It was suggested not to refer to temperature classes defined in a standard in the definition. JRC agrees with this.

Minutes and summary Ecodesign for Commercial Refrigeration 1 st Technical Working Group meeting JRC-IPTS, Seville, 23 April 2013, 9h00-17h30 DISCLAIMER: The views expressed are purely those of the writer and are intended to reflect the discussions at the above mentioned meeting. They may not in any circumstances be regarded as stating an official position of the European Commission. 1. Scope and definition The definition proposed in the background document was in general approved. A new version of the scope will be distributed to the stakeholders within a few weeks, for comments and clarification, as it is the objective to settle the agreement on the scope and not discuss the issue in the next TWG meeting, except for smaller clarification questions. It has been suggested to restrict the scope to food products. If stakeholders believe the widening of the scope to also include non-food products may result in an additional burden, they are requested to signal it and provide the technical arguments for it, as otherwise the wider scope will be kept. An expert suggested that a distinction shall be made between direct and indirect sale. Direct sale involves no action, while indirect sale involves some manipulation by a professional (milk added to coffee, ice cream put in waffle cone). Indirect sale may be regarded as professional refrigeration. This distinction and proposal was not supported by other experts, the generic interpretation being that products on display are commercial refrigeration, directly or indirectly. Experts that believe that there are technical differences in the appliances due to this distinction are kindly requested to provide data to support it. It was suggested not to refer to temperature classes defined in a standard in the definition. JRC agrees with this. A stakeholder mentioned that walk-in cold rooms, in e.g. wholesale supermarkets, are currently not included in Lot 12 Commercial Refrigeration, although some of them fulfil the definition. In general, cold rooms are included in Lot 1 Professional Refrigeration. However, there is still a lot of work to do for this product group in Lot 1. 1

A stakeholder suggested moving professional service cabinets, currently under Lot 1 Professional Refrigeration, to Lot 12 Commercial Refrigeration because they are very similar in construction. However, there is a difference in insulation, material choice, lighting, double/triple layer glass doors, defrost systems, as well as price. Aiming at the same MEPS for professional service cabinets and commercial display cabinets could lead to an inconsistent value setting. Also ambient temperatures during testing differ for Lot 1 and Lot 12 appliances: at the moment, it is suggested that Lot 1 appliances are measured in climate class 4, while Lot 12 appliances would be tested according to climate class 3. It is desirable that the scope and definition refer to clear distinctive physical features/characteristics of the product. Using intended use as a parameter is less clear. This is not always possible for instance in many cases, devices that look like household or professional from a physical-technical point of view become commercial through image/stickers/labels attached to them. There was general agreement on the product groups to be excluded from the scope: water dispensers, ice cream makers, ice makers, minibars (under domestic refrigeration regulation), wine cellars (sold for domestic and for commercial use). Chest freezers would be under the scope of domestic refrigeration regulation this has been discussed in Lot 1 meetings. Gelato freezers are in principle included, as they fulfil the definition of commercial refrigeration. However, this product group is apparently known for not having in the past worked intensively on energy efficiency nor on recording data on this issue. Standardisation will be an issue, as the preparation of standards takes time, and not all commercial appliances covered in the scope are currently referred to in a standard. Gelato freezers are an example of this. 2. Standards and legislation Most of the relevant international legislation and standards were correctly spotted in the background document distributed. Some of the international deserve closer attention, as they include approaches that could be of use for an EU regulation on commercial refrigeration. Stakeholders were encouraged to use the opportunity of written comments by 24 May to explain their experience with barriers to ecodesign created by any current legislation or standards, and to explain specific cases of redundancy that they may have detected. This would help design a diagram explaining the requirements and barriers of current standards/legislation. Safety The main legislation related to safety issues are the Pressure Equipment Directive (PED) and the ATEX Directive for explosive atmospheres. The PED has to be followed with any device working under pressure. Under the ATEX Directive a load of harmonised standards can be found. The most 2

important standard is EN 378 which is harmonized with the European Pressure Equipment Directive. ISO 5149 is not relevant for these products, as it is unusable. Even if it is under review and it will be voted, it is expected that it will not be adopted in any of the Member States. More relevant for Europe is the EN378 standard. EN 378 is a horizontal standard and governs the selection of refrigerant fluids and permitted charge sizes for given applications (in terms of the occupancy classification) and details criteria for equipment design, construction/installation, and maintenance. If EN378 is applied, all safety requirements are expectedly met. If the charge of a flammable refrigerant exceeds 150g in an appliance, usually a risk assessment is carried out following EN 1127. When the charge exceeds 150g, the manufacturer is responsible for the safety measures, and it cannot be certified by a 3 rd party. Taking into account that applying standards is voluntary, different manufacturers can apply different conditions, whether or not safe, which can create unfair competition. Technically, there is no problem to produce safe appliances. However, some stakeholders claim that safety standards are blocking manufacturers that want to have the same level playing field. Other argue that by applying sometimes unnecessary safety measures (in cases of duplicity), production price increases. Another rise in cost can be that ATEX compliant components, e.g. electrical boards, are usually 5 times more expensive than components used for non-explosive/non-flammable refrigerants. In principle, there is no safety barrier. Also a lack of available components on the market can increase the consumer price significantly. This is the case for CO2-technology using appliances. The problem with CO2 is the cost rather than the safety, especially for smaller appliances. In general, legislation, standards, technical and economic factors have to be considered together. In the PED for example, a pressure relief device is required which in the case of plug-in appliances is unnecessary (as confirmed by TNO studies). According to PED, the high pressure side should be able to stand 3 times the actual working pressure. For a R134 appliance, this means 3x30bar = 90bar, while for CO2 appliances this means 3x140bar = 420 bar. Vending machine manufacturers describe themselves as followers in this business and normally would buy the cooling system as a module to be inserted in the machines. Concerning safety, they have the additional concern of mechanical moving parts and electronics that can produce sparks that can ignite the flammable refrigerant in case of leaks. For the moment, no vending machines are produced with hydrocarbons. In case they were to use hydrocarbons as refrigerants, vending machines can use a double-trap safety system applying sniffers and a venting system. Technically it is possible, but the cost is high, limiting these machines to enter the market. Most vending machine producers are moving to CO2, and not to hydrocarbons. CO2 appliances are widely used in supermarkets, especially in Northern Europe, and safety problems were encountered only in the early years. Some accidents were mentioned, but these are probably not to be attributed to the specific technology, as also for other more established technologies 3

accidents have been reported. Installations working on hydrocarbons with a charge bigger than 150g are installed in supermarkets and no (safety) problems have been reported so far. Once the limit of 150g has been passed, a grey zone is entered and no limit on quantity is defined. For the risk analysis, following EN 1127, different routes can be followed. This implies that clearer safety standards are necessary, especially for plug-in cabinets. As EN 378 is currently under revision by CENELEC (TC182), a mandate could be send by the European Commission to take these concerns into account. Concerns about the duration of such a process are expressed. It is also mentioned that term natural refrigerants is misleading and non-technical. It is suggested to use non-fluorinated refrigerants instead. Energy use A number of cabinet types are currently not covered by ISO 23953: corners, angles, open-ends, back loading, gelato freezers. The potentials of presence-detection devices are not currently considered. Moreover, such energy saving devices are mainly used for beverage coolers and not for machines containing food, where the temperature cannot rise as it can with beverages. The EVA-EMP for vending machines which will be sent to CENELEC intends to include such energy-saving devices. In principle these devices could be tested also in ISO 23953, but are not included in the standard. Vending could be categorized in three classes regarding the products they sell: beverages (usually brand owned), snacks (like chocolate bars) and food (temperature cannot rise above a certain temperature). The Japanese measurements standards JIS B 8561 shall be scrutinised (if not in Japanese Language only) for information about vending machines on lighting. It is also mentioned that a Japanese standard or a non-european standards in general, are not applicable as they are generally designed for a different scope, especially for vending machines. The ASHRAE standard for example only deals with can & bottle vending machines. It has to be also checked if ISO 23953 also covers appliances not intended for food and beverages. It is not sure if drawers in display cabinets are taken into account in the ISO 23953:2012 version. Beverage coolers are in principle covered by the ISO 23953 standards, but they (should) have different requirements compared to vertical display cabinets. These appliances are usually tested with a brand-specific measurement protocol, which is usually stricter regarding energy efficiency than ISO 23953, and takes into account specific parameters (e.g. presence detectors). An example of an extra requirement can be a maximum time to pull-down the temperature quickly, for example after refilling. Such a pull-down temperature/time is not mentioned in ISO 23953. Canada, US and Korea have a pull-down test included in their standards. Bottle coolers usually also fail the ISO 23953 relating to the working temperature. This is mainly attributed to the filling protocol which does not represent the actual situation if the cooler is filled with bottles rather than with filler packs. These filler packs obstruct the cool air circulation more than bottles do, preventing the machine to maintain its temperature. 4

Testing at climate class 3 is relevant. For domestic refrigerators, energy consumption tests are done at 25 C to establish a common playing field, but they are additionally classified as tropical, central and northern climates. There is a difference between the performance requirement and testing the appliance at certain ambient conditions. 3. Markets The focus of the ecodesign preparatory work is on energy performance, and not primarily on market data. However, the JRC explained that the objective of collecting updated information on markets has as sole purpose establishing a solid foundation for the estimates of energy saving scenarios that are part of the Impact Assessment. If only old data were used, this may result in proposing unrealistic ecodesign criteria, and the need to relaunch market data collection at a later phase, which will be inefficient. The Eurovent certification program covers around 50% of the market, while the association covers around 80% of the manufacturers. EVA has provided market data. Some of the stakeholders have offered to share with the JRC data on the stock of their fleet, including expected growth, refrigerant used and energy-saving options present or not. Current and past data can be provided, but projections of sales prove difficult, as the market situation has changed in the last few years and it is more difficult to forecast. A consistency check could be done by comparing the market data with an SKM Enviros study about refrigerants which should be available through the website of DG Clima. Another study by Ecole de Mines together with EPEE about refrigerant consumption, emissions, banks, etc. could also serve this purpose. 4. Energy The point of departure of the JRC work is the energy efficiency formulae proposal of the Impact Assessment of Wuppertal Institute (2010). The JRC is at this stage in the task of understanding the details of the parameters part of these formulae, a process that will be done as systematically and transparently as possible, and will likely involve the stakeholders through a questionnaire in late summer 2013. 5

The key objective of the analysis and reassessment of the existing formulae is to find an appropriate breakdown level for the product groups addressed. Energy formulae are to be kept as generic and simple as possible, but contain key elements that enable to adjust it for product group specificities that affect significantly their energy consumption compared to other product groups. In the proposal of Wuppertal Institute, and in other regulations, this has been done by use of parameters, which take different values depending on the product group characteristics. The energy formulae have to contain as a minimum the 2-3 essential parameters that make the lion's share of energy consumption in a cabinet. As regards labelling, there is no strong push or opposition from the experts to introduce it. Priority will obviously be to the product groups where the largest energy efficiency benefits are expected. JRC presented the data collected so far. There is a large discrepancy between existing updated data and the older database that was used by WI to propose its formula. Stakeholders are encouraged to provide (if needed anonymised) recent energy use data, so the JRC is able to prepare a reliable energy efficiency formula. Some of Eurovent members provided some possible explanations for the above mentioned discrepancy. According to some of them, Eurovent certification database contains often best available cabinets technically available at the moment (this is in contrast to the comments received in this respect by some of the manufacturers, indicating that the products tested are average products). There can also be a difference between cabinets with and without night blinds. Another possible explanation is that the energy efficiency improved, but the TDA also increased, resulting in no net improvement of the 1997 average values and the database average values. Another explanation factor is that the data in the Eurovent Certification database are all with unlightened shelves. Data on differences for cabinets with and without night blinds can be obtained. The conclusion of this issue is that a rational explanation to the mentioned discrepancies needs still to be provided by Eurovent if the proposed average used for certification is to be used. It is suggested to not make a difference in energy performance between open and closed cabinets. Only the application and the temperature class should be taken into account. Energy efficiency can also be market-dependent. Experts mention that in Scandinavia no shelf lighting is used, and indirect lighting from the store is used, while opposite practices are common in Southern Europe. In Northern Europe it is also more common to have closed appliances, a trend which is expanding to the rest of Europe. The Wuppertal formula with lighting is apparently not correct. Lighting contributes in reality to the energy consumption more than proposed in the formula. In the Wuppertal formula, canopy lighting is taken into account, but not shelf lighting which is more energy demanding. This will be scrutinised. A direct comparison with plug-in and remote cabinets may be difficult as they usually have a different configuration. Plug-ins are usually smaller cabinets. Background on the data and thresholds set for the ECA scheme will be provided by some of the stakeholders. Some experts indicate that plug-ins are usually less energy efficient as they use (auto-) defrost mechanisms and water removal 6

systems which could add up to 1.5kW/(day.m²). Also the use of a capillary tube as expansion valve and the use of a hermetic compressor (compared to a semi-hermetic one) would result in less energy efficiency for plug-in appliances. The working temperature in the WI formula is defined as the average temperature of all the M- packages during the test. It is suggested that it is better to use minimum and maximum temperatures. It is important that an energy formula is grounded in empirical evidence. To get a whole spread of the market, it could be good to compare also with data from Australia, Canada and USA. Data for bottle coolers could be obtained from USA's Energy Star and from Australia. Some benchmarking studies indicate big differences in energy efficiency of vending machines between EU, Australia, and USA. This is mainly attributed to a difference in user behaviour resulting in a different stock, and different appliances on the market. EVA wonders how this difference can be, as many machines are produced by the same manufacturers, and sold in different regions in the world. The energy consumption should intrinsically be the same. EVA has volunteered to check the Wuppertal formula for vending machines against their database of energy consumption and the EVA-EMP classification. EVA indicates that vending machines have experienced an improvement in energy efficiency was done over the years data needs however to be provided to support this statement. The differences between TEC vs. TDA should be analysed carefully. TDA is apparently a sensitive variable, and is easy to manipulate by e.g. fixing shelves. TEC/m would be better. Most stakeholders however agree with a TEC/TDA value. The MEPS proposals shall not just be an energy efficiency formula, as other options (recommendations, mandatory elements, etc.) are possible. There is a e.g. lot of improvement potential in fitting the different parts together, especially for very inefficient appliances. As indicated in the background document, there is a lot of info available on energy efficiency from Carbon Trust's Refrigeration Road Map. Some manufacturers mention that for them it is in general easier to work on adjustment of energy efficiency on devices that use standard refrigerants, as they have more control and knowledge of the variables. On new refrigerant, priority is to build up expertise on stability. Experts mention that many retailers will invest in energy saving options/appliances only if the payback time is less than 2 years, usually 4months-1yr. Labelling EVA mentions that labelling on vending machines could be misleading. The borderline between its use as a technical tool and a commercial tool is not clear to some stakeholders. It is important to set a reliable benchmark, and allow a third party to certify consumption. Such third-party certification are difficult to implement. Labelling could be done for standard products, but could be difficult for custom-designed products. 7

Denmark and its industry are in favour of labelling of the products in general and in particular for plug-in cabinets. Data has been delivered from the Swedish Energy Agency. Plug-in appliances can be classified as machinery and labelling would be appropriate. Remote display cabinets however are not machinery, but their energy consumption is affected by the whole refrigeration system. It would be better to label the complete system than the single appliances. However, a system with good (labelled) cabinets would only make the system better. Labelling is a good tool to begin teaching customers (retailers) about incorporating energy efficiency as a decision-making criterion, also for the remote part of a system. A cost efficiency breakdown should be done. There should be data available on this from the USA/Australia/Canada, as in some of these labelling is mandatory. 5. Refrigerants The initial approach of the JRC is to ensure alignment with the objectives of the F-gas regulation under current debate/negotiation (some outcome is expected for the summer 2013). In principle, there should be no need to intervene and include any specific clause on the use of refrigerants in the ecodesign regulation. If needed because the F-gas regulation is not deemed to sufficiently address the specificity of commercial refrigeration, one could device a system in the ecodesign regulation to reward efficient, low GWP refrigerants. This could be a hard system (e.g. a bonus-malus element in the efficiency formula)or a softer device (labelling), in any case taking a life-cycle consideration (e.g. not to push low GWP refrigerants in climatic areas where these are not efficient). There seems to be abundant but dispersed information about refrigerant efficiency, especially from beverage and food manufacturers, and retailers. A Total Equivalent Warming Impact (TEWI) approach is in general used, as compared to a blunt comparison of GWP. This means just acknowledging that the largest GWP impacts and savings are on the use phase, and relate to the efficiency in heat transfer of the refrigerant, and not that much to their intrinsic GWP. Subcritical CO2 systems work fine in warmer climates. If they are less energy efficient in warmer climates, promoting low GWP appliances at all costs seems not beneficial on a life-cycle perspective both environmentally and financially, as the end-user would bear both loads. Data on efficiency for HFOs in supermarkets will be provided by some of the manufacturers. Some large food and beverage manufacturers have large programmes to replace refrigerants to nonfluorinated refrigerants. Some of them have data comparing efficiency of the same cabinet with different gases, and will be provided to the JRC. 8

For a remote cabinet, the impact of the refrigerant exists but it is less important than for plugins, as many other aspects have a role in the overall performance: dimensioning, piping, distances, efficiency of the external compressor units, etc. The choice of the refrigerant in remote cabinets is a decision of the retail system designer, not of the cabinet manufacturer- however, a close communication is expected between them, as this helps optimise the systems (both the cabinet and the non-cabinet parts). It was proposed that there should ideally be freedom to the manufacturer to choose the refrigerant, once the guiding principles and objectives of the F-gas regulation are clear. It will then be up to the manufacturers to adapt and balance design to follow this, and incorporate cost and efficiency elements. 6. End-of-life JRC-IES presented a wish list of possible options (e.g. dematerialisation, clearer labelling of critical elements (F-gas for refrigeration and foaming), modularity of appliances and easier dismantling). The JRC will work intensively before/during the summer, in close cooperation with retrofitters/recyclers, to select on 2-3 of these options at detailed level. Priority is to be given to measures that facilitate legislative compliance (mainly WEEE and RoHS), followed by material recovery. The manufacturers see a priori no major difficulties in the suggestions provided. Some of them (labelling of both refrigerant and foaming gas) seem very easy to implement. Moreover, labelling of the refrigerant gas used for the cooling system is already mandatory according to the F- gas regulation. Additional questions to the stakeholders on the EoW can be expected included in the next questionnaire(s) to be circulated by the JRC before/after the summer. There is a clear mismatch of the characteristics of the appliances currently being recycled (which arrive currently to recycling plants) and the ones currently commercialised (newer, no CFCs, clear labelling, etc.). Any measure shall address the current new products, not the older ones of characteristics which in 5 years will no longer be received. JRC will investigate further WEEE interpretation and Waste Shipment Regulation compliance in the MS for commercial refrigeration. The objective is to get first indications of the potential impact of requirements that affect the EoL. The situation is at the moment: o WEEE: Currently (WEEE 2002), those appliances not similar to household refrigeration appliances can be categorised as non-weee. However, most Member States consider vending machines and bottle coolers as (close to) household refrigeration and treat them accordingly. Following WEEE recast (2012), from 15 August 2018 all commercial appliances will clearly be under the scope of WEEE. Ecodesign may have direct effects, e.g. more efficient removal of hazardous components (batteries, electronic parts): 9

o WSR: Commercial refrigeration has a high percentage of reuse and retrofitting, often involving trade. Trade out of the EU can be of concern, and the compliance with the WSR has to be checked. This has a priori only indirect effects on an ecodesign regulation, and concerns mainly the impact assessment (a.o. to what extent can the new ecodesign measure result in more /less trade within/out of the EU). 10