TL: REFRIGERATORS, ENERGY & CLIMATE: MANDATORY ENERGY EFFICIENCY STANDARDS FOR DOMESTIC REFRIGERATION UNITS IN THE EUROPEAN UNION [* NOTE 85 PAGES LONG!] SO: PAUL WAIDE AND HORACE HERRING, GREENPEACE INTERNATIONAL, (GP) DT: DECEMBER, 1993 [PLEASE NOTE, DUE TO THE LIMITATIONS OF TEXT-ONLY FORMATTING, NUMEROUS GRAPHS, AND THE ENTIRE APPENDIX ARE NOT AVAILABLE IN THIS VERSION OF THE REPORT. FOR THE FULL VERSION, PLEASE CONTACT GREENPEACE INTERNATIONAL] TABLE OF CONTENTS OVERVIEW EXECUTIVE SUMMARY TECHNICAL AND ECONOMIC ANALYSIS: SECTION 1: MANUFACTURERS AND MARKETS 1.1 THE EUROPEAN MARKET 1.1.1 PRODUCT CATEGORIES 1.1.2 NATIONAL MARKETS 1.1.3 SALES TRENDS 1.1.4 OWNERSHIP AND MARKET MATURITY 1.2 PRODUCTION WITHIN THE EU 1.2.1 LEADING EU MANUFACTURERS 1.2.2 BRAND NAMES 1.2.3 RECENT MERGERS AND TAKEOVERS IN EUROPE 1.2.4 MANUFACTURER MARKET SHARES BY NATION 1.2.5 NATIONAL PRODUCTION, EXPORTS AND IMPORTS SECTION 2: ENERGY MEASUREMENTS AND DATA 2.1 MEASURING REFRIGERATOR ENERGY CONSUMPTION 2.2 ENERGY DATA FROM THE EUROPEAN REFRIGERATION MARKET 2.3 MANUFACTURER REPORTING OF ENERGY CONSUMPTION SECTION 3: MEASURES OF APPLIANCE ENERGY EFFICIENCY 3.1 ENERGY EFFICIENCY INDEX 3.2 PRODUCT CATEGORIES SECTION 4: REFRIGERATOR DESIGN ISSUES 4.1 DESIGN CONCEPTS 4.1.1 COOLING SYSTEM OPERATING PRINCIPLE 4.1.2 CABINET DESIGN AND THERMAL LOADS 4.1.3 DEFROSTING SYSTEMS 4.1.4 IMPACT OF OZONE DEPLETION ON CHOICES OF REFRIGERANT AND FOAM BLOWING AGENTS 4.1.5 MIXED FRIDGES AND FREEZERS 4.2 THE TECHNICAL POTENTIAL FOR IMPROVING REFRIGERATOR ENERGY EFFICIENCY 4.2.1 IMPROVING THE EFFICIENCY OF THE COOLING SYSTEM 4.2.2 IMPROVING THE EFFICIENCY OF REFRIGERATOR INSULATION 4.2.3 VACUUM INSULATING PANELS (VIPS) 4.2.4 SIMULATED REFRIGERATOR ENERGY CONSUMPTION USING VIPS SECTION 5: SHORT-TERM ENERGY STANDARDS 5.1 ENERGY REFERENCE LINES 5.1.1 GEA REFERENCE LINES 5.1.2 COMPARISON OF GEA AND ECO-LABELLING STUDY ENERGY REFERENCE LINES 5.2 DISTRIBUTION OF REFRIGERATOR ENERGY EFFICIENCY 5.2.1 DISTRIBUTION OF ENERGY EFFICIENCY BY NATIONAL MARKET 5.2.2 DISTRIBUTION OF ENERGY EFFICIENCY BY PRODUCT CATEGORY 5.2.3 DISTRIBUTION OF ENERGY EFFICIENCY BY MANUFACTURER 5.3 DEFINITION OF SHORT-TERM STANDARDS 5.3.1 CONVENTIONAL MODELS 5.3.2 NO FROST MODELS 5.4 IMPACT OF SHORT-TERM ENERGY EFFICIENCY STANDARDS ON NATIONAL REFRIGERATION MARKETS SECTION 6: LONGER-TERM ENERGY STANDARDS 6.1 METHODOLOGY 6.2 CALCULATING COSTS 6.3 THE GEA DESIGN OPTIONS 6.4 DESCRIPTION OF THE GEA BASE-CASE MODELS 6.5 RESULTS OF THE ENGINEERING ANALYSIS 6.6 THE GEA RECOMMENDATION FOR LONG-TERM MINIMUM EFFICIENCY STANDARD LINES 6.7 SOCIETAL COSTS 6.8 LEAST LIFE CYCLE COSTS FOR VACUUM INSULATION PANEL DESIGNS SECTION 7: GREENPEACE PROPOSALS FOR MINIMUM EFFICIENCY STANDARDS SECTION 8: STANDARDS ENFORCEMENT 8.1 ENFORCEMENT ADMINISTRATION 8.2 COST OF ENFORCEMENT ADMINISTRATION SECTION 9: IMPACT OF ENERGY EFFICIENCY STANDARDS 9.1 METHODOLOGY 9.2 THE BASE-CASE SCENARIO AND HISTORICAL EFFICIENCY IMPROVEMENTS 9.3 RESULTS SECTION 10: THE CONTRIBUTION OF HFCS TO THE GLOBAL WARMING COMMITMENT OF A TYPICAL REFRIGERATOR REFERENCES APPENDIX: SCENARIO FORECASTS ---------------------------------------------------------------- OVERVIEW This study presents a technical, economic and political argument for the rapid introduction of minimum energy efficiency standards for domestic refrigerators in the European Union. The study is divided into two parts: 1) the Executive Summary, 2) the main technical and economic analysis. The Executive Summary delivers a condensed version of the main study and a critique of the minimum energy efficiency standards proposals of the European Commission. Greenpeace's own standards proposals are advanced. The main report contains the detailed technical and economic analysis upon which the political arguments of the Executive Summary are based. Stylistic Note: The word `refrigerator' is used throughout to refer to fridges (refrigerators), freezers and fridge-freezers collectively. Executive Summary The Directorate General for Energy of the Commission of the European Communities (DG XVII) are about to issue a draft directive proposing mandatory minimum energy efficiency standards for refrigerators, freezers and fridge-freezer units sold within the EU. The proposed standards are intended to reduce the electricity consumed by the stock of new models by 10% from the average consumption of models sold in 1992. The standard would come into force at the beginning of 1997. This report presents a critical analysis of the proposed standards based upon the results presented herein. Alternative standards are proposed which are more consistent with the technical and economic evidence and the need to meet environmental and societal goals. The main conclusion is that the EU proposed standards are too weak and that it is likely that energy efficiency performance would improve faster than targeted simply by the effect of existing market pressures. Should the proposed standards become legislation in their present form there is a real possibility that some member states will seek to impose more meaningful standards unilaterally. Summary Critique of the Proposed EU Standard: ù The proposal of an average -10% standard is far too weak and has no technical or economic basis. ù The delay in imposition of the standard until 1997 is not justified as a -10% standard could reasonably be met in less than a year. ù The proposal that consideration of tougher standards based upon an economic and technical analysis be delayed until 1997 is without justification. The necessary work has already been performed by the Commissions own research team, the Group for Efficient Appliances (GEA 1993), but their recommendations have been ignored. ù The proposed standard constitutes a weaker efficiency improvement than would be anticipated even if no standards were enforced. For this reason there is likely to be little or no environmental or economic benefit from its introduction. ù The maximum electricity savings resulting from the Commissions proposal would have the effect of preventing cumulative emissions of only 5 million tonnes of carbon dioxide (CO2) by the year 2000, and perhaps 39 million tonnes by 2010. The Commission's proposal may result in no electricity savings or emissions reductions at all. Standards proposed by Greenpeace, however, could prevent between 25 and 27 million tonnes of CO2 emissions by the year 2000 and between 170 and 254 million tonnes by 2010. Ultimately the Greenpeace standards process would probably lead to a 5 to 6 fold increase in the efficiency of refrigerator models sold in 2000/2 over the 1992 models. ù The proposal ignores the precedent set in the United States where standards were successfully introduced even though they required almost all models on the market in 1990 to be removed from sale by 1993. Efficiency improvements of between 25% and 30% were obtained for less than a 1% purchase price increase borne by the consumer (Turiel 1993). An analysis by the Lawrence Berkeley Laboratories on the impact of the 1993 standards indicated they "would not diminish the profitability of the industry" (McMahon 1992). ù The draft Directive contains no reference to the need to ensure that domestic refrigerators do not employ ozone-destroying chemicals, such as CFCs and HCFCs, or new classes of greenhouse gases, such as HFCs. A simple analysis of the probable contributions of HFCs to the total equivalent warming commitment over the lifespan of a typical European refrigerator are presented herein. It is shown that using HFCs increases the total equivalent warming impact (TEWI) of a typical refrigerator by 15%. ù The enforcement measures suggested in the proposal are wholly inadequate; in particular, the reliance upon self-policing by manufacturers. This conclusion is supported through independent test-data collected by consumers organisations which indicates that on average manufacturers understate the true energy consumption of their models. ù The lack of a long-term economic and environmental appraisal group within the EU, which would continually review and propose updated standards on a regular basis is a major weakness in the current draft Directive, which will significantly delay efficiency improvements. ù The wording of the proposed standard is imprecise in certain important respects which may encourage some manufacturers to try and exploit loop-holes. Summary of Greenpeace Proposals: ù The introduction of a standard set at 15% below the 1992 average for all models by January 1st 1995 should be announced by January 1st 1994. ù The introduction of minimum energy efficiency standards for January 1st 1997 equivalent to a reduction in the electricity consumption of new refrigeration units of between 44% and 50% (depending on category) of the electricity consumed by average 1992 models. Notification of the standards should be made by January 1st 1994. These standards are effectively the same as the `long-term standards' proposed by the GEA (1993). The date for introduction has been advanced from 1999 to 1997 mainly on the basis of the successful US experience, wherein all models on the market were replaced within a three year period without loss of model choice, appreciable increase in price (Turiel 1993), or reduction of sales. ù The standards should be under continual review with fresh targets set every three years, as in the US. The assessment of the standards should be conducted by a cross-national, politically-independent, research team, such as the GEA, whose findings would be binding. Standards should not be assessed by non-technical personnel in the Commission. This would require the EU to establish and fund a Technical Committee, which would have responsibility for proposing new standards every 3-4 years. ù New standards should be announced at the start of 1997 to come in effect by the year 2000 to 2002. ù Clear accounting principles should be adhered to in setting the standards. As a bare minimum these should aim to set the standards no higher than is implied by the Least Life Cycle Cost of the appliance to the consumer. Preferably the standards would be based upon the Least Societal Cost which would incorporate an assessment of the value of preventing the emission of pollutants. ù The standards should not be policed by the manufacturers but by an independent specialist agency established by the Commission. Every manufacturer should be required to submit at least 3 models for testing and the average electricity consumption should be quoted. Additionally the agency should have the power for random testing and if a model is found to consume electricity 15% in excess of the quoted value a further 10 models should be tested for compliance. If the average of the 10 models is found to exceed the standard by 5% then the model will have failed the test. If necessary this agency could sub-contract testing to the facilities operated by national consumer groups. ù If a manufacturer is found to be in breach of the minimum standard at the initial testing phase the appliance should be barred from sale in the EU. If however, a manufacturer is found guilty of attempting to deceive the testers at the initial testing phase the models should be withdrawn and a financial penalty incurred. Introduction The Energy Directorate of the European Commission is about to issue a draft directive proposing mandatory energy efficiency standards for domestic refrigeration units sold within the European Union. The impetus for the legislation arose from the Dutch government proposal to introduce unilateral minimum energy efficiency standards for refrigeration units that was challenged by the Commission on grounds that the standards would contravene the free-trade terms of the single market. A compromise solution was struck, known as the `Dutch notification', by which the Commission were obliged to produce their own proposals for union wide mandatory standards within a year of the Dutch notification issued on January 8th 1992. The formal announcement of the draft proposal is still awaited, more than 10 months later. Why Introduce Mandatory Standards ? The European Union have committed themselves to the stabilisation of CO2 emissions at 1990 levels by the year 2000. Apart from the carbon/energy tax (which is currently stalled because of British opposition) the Directives concerning appliance energy usage are the only measures proposed which openly address this target across the whole Union. It is generally recognised that the imposition of minimum energy efficiency standards results in more rapid and quantifiable energy efficiency improvements than the use of fiscal incentives. Failures occur in an unregulated market because consumers do not have access to much of the knowledge necessary to take optimal financial and ethical decisions. The introduction of energy and eco-labels can (if done properly!) assist in bridging this knowledge gap. However labelling measures are insufficient on their own, in part because the labels tend to only give information which allow relative comparisons between models currently on sale. Minimum mandatory energy efficiency standards offer the following advantages over fiscal incentives: ù they set minimum defined performance improvements within predetermined time scales which enables a more reliable estimation of the outcome, ù they can be based on reliable assessments of what is technically and economically reasonable for manufacturers to achieve at a cost beneficial to the consumer, ù they encourage knowledge pooling between experts and manufacturers and remove the need for an unrealistic level of technical expertise among consumers, ù they can encourage manufacturers to pool RD&D costs thereby lowering the net cost of technical improvement as experienced in a unregulated market, ù they can force product development to occur at a faster pace than would otherwise have occurred, leading to enhanced manufacturer competitiveness. WHY LEGISLATE ON REFRIGERATION? In energy terms refrigeration units are the dominant domestic appliance. Across the European Union refrigerators, freezers and fridge-freezers account for 111.2 Terrawatt-hours of electricity use per year or 24% of all domestic electricity use (6% of total electricity demand). This level of consumption requires 16.9 GW of generation capacity (i.e. about 17 large power stations) and is estimated to result in emissions of 61 million tonnes of CO2, 815 thousand tonnes of SO2, 430 thousand tonnes of NOx and total greenhouse gases with a global warming potential (GWP) equivalent to 73 million tonnes of CO2 (assessed over a 50 year integration) every year. Of all the domestic appliances refrigeration units have been singled out for treatment first because they offer the greatest scope for energy reduction through the imposition of standards. In 1992 the most energy efficient models available on the European market were between 25% and 65% more efficient than the average models of their categories. Furthermore, by comparison with other large appliances, it is relatively simple to define satisfactory energy efficiency measures which have no detrimental impacts upon other aspects of the product. It is important to remember that European efficiency standards have major implications outside Europe as well as inside. Western Europe is the largest market for refrigeration appliances in the World and its manufacturers have a major export and technology transfer potential. European Free Trade Association (EFTA) countries such as Austria, Norway, Sweden and Switzerland will almost certainly follow EU standards as they move towards becoming full members of the Union. A number of east European countries such as Poland, Hungary and the Czech Republic will also tend to adhere to EU standards for similar reasons. The implications of European standards for energy efficiency are especially important for developing countries where huge consumer markets are demanding increased access to consumer appliances such as refrigerators. For example, China, which manufactured no refrigerators in 1980, made more than 8 million in 1990 and has a stated aim to supply a refrigerator to each household. India has projected that ownership of refrigerators will grow from 4% of households in 1992 to 60% by 2015. These countries will require an enormous growth in electrification just to supply domestic refrigeration demand. This will greatly drain scarce capital and exacerbate local air pollution and global warming problems if the models they adopt are inefficient. BACKGROUND OF THE EU DRAFT DIRECTIVE Summary of the EU Proposed Standard The Commission have proposed mandatory minimum energy efficiency standards in accordance with the -10% standard of Table 5.4 (see Section 5 of the technical and economic analysis) for refrigeration unit categories R1, R2, R3, R4 and F2, and in accordance with the -15% standard of Table 5.4 for category F1. These standards are to be implemented by 1997. Under Article 14 of the draft Directive it is stated that a proposal for a more demanding set of standards based upon a technical and economic appraisal be submitted in 1997. The proposal may then require a further 2 years before it is passed as legislation, which would delay the announcement of tougher standards until 1999 and prevent them coming into effect until 2002 at the earliest. The submission of a proposal for tougher standards is to coincide with the completion of an assessment of the implementation of the Directive and "the results obtained and foreseen" (It is noted that the date for submission of the assessment is the same date as the -10% standard is due to be implemented!). Assessment of the appliance category is left up to the manufacturers and the standards only apply to the listed appliance categories. Unfortunately, the definition of whether a given appliance belongs to a given category is open-ended and vague. Article 3 states: "Refrigeration appliances shall be considered to belong to the same type, referred to in this Directive as "appliance type", if they are produced by the same manufacturer (or under licence by a different manufacturer) and differ only in aspects which do not significantly affect their energy consumption in use in any way" The looseness of definition might encourage manufacturers to modify some features of their appliances so they could argue they are outside the standard categories, e.g. a refrigerator with a freezer compartment operating at -16oC. The proposed enforcement of the standards is left to self policing by manufacturers, while the penalty for non-compliance is removal of the offending model from sale. THE GEA RECOMMENDATIONS The standards proposed by DG-XVII represents a substantial climb-down from the recommendations of the Group for Efficient Appliances (GEA) who were contracted by the Commission to assess the appropriate basis for mandatory standards. The Commission have dodged the recommendation of a tougher longer-term standard and have set the proposed standard at a level which is seen to be weaker than any of the recommended short-term standards. The GEA who comprise researchers from French, Dutch, Danish and Portuguese governmental agencies conducted a detailed technical and economic assessment of the European refrigeration market. They proposed that short-term mandatory energy efficiency standards should be implemented as in interim measure before introducing tougher longer-term standards. The toughness of the standards were linked to the date of introduction in acknowledgement that efficiency will continue to improve through technological advances stimulated by market pressures in the intervening years. The long-term standards are set at a value which the GEA estimated corresponded to the lowest overall cost to the consumer, the so-called Least Life-Cycle Cost. This corresponds to the point at which extra efficiency measures would add more to the purchase cost of the appliance than they would benefit the consumer in reduced electricity consumption over the lifetime of the appliance. In setting the long-term standards the GEA only considered proven efficiency technologies. The long-term standards they proposed would result in energy efficiency improvements of between 38% and 55% over the 1992 models depending on the category of appliance. The GEA proposed the following standards scenarios (the values are relative to the average energy consumption of models on the market in 1992): Scenario 1: Implementation of a -10% Standard in 1995 and the Long-Term standard in 1999. Scenario 2: Implementation of a -15% Standard in 1995 and the Long-Term Standard in 1999. Scenario 3: Implementation of a -15% Standard in 1997 and the Long-Term Standard in 2001. The details of the standards approach and the appliance categorisations proposed by the GEA and accepted by the Commission and Greenpeace are described and explained in the main technical and economic analysis (Sections 3,5 & 6). It is clear that the -10% standard to be implemented by 1997 as proposed by the Commission significantly fails to meet the GEA recommendations. TECHNICAL POTENTIAL FOR IMPROVED EFFICIENCY The energy required by a refrigeration unit is determined by the size of the total thermal load and the efficiency of the cooling system which satisfies it. The most cost effective way to reduce energy consumption is to reduce the thermal load (and hence electrical load) through better insulation of the cabinet. According to Benson and Potter (1992): "Contrary to popular opinion, the energy use of a household refrigerator is not closely related to the frequency of door openings or the amount of food kept inside. Rather even with improvements in insulation over the past 20 years, between 70% and 95% of the electrical load, may still be attributed to the thermal performance of the insulated shell. The remaining load is caused by gaskets, which also have great improvement potential, and from (automatic) defrost heaters and antisweat heaters (which are not so common in European models)." In the engineering analysis conducted by the GEA to assess the longer-term standards they considered the effect of increasing the cabinet insulation thickness, improving the door seals, using more efficient compressors, and using better evaporators and condensers. They found that using tried and tested, readily available technologies that the Least Life Cycle cost coincided with energy efficiency improvements of between 38% and 55%. These results were obtained assuming only conventional foam blown insulation of limited thickness so as not to seriously effect the storage space or size of the appliance. Payback periods of between 3 and 4 years were obtained. Traditionally the cabinet insulation has been provided by blown foams such as polyurethane, however recently evacuated panel technologies have been developed which offer the opportunity for greatly improved energy performances and a better all round product for the consumer. In the United States a manufacturer of vacuum panels (Owens-Corning) has teamed-up with a refrigerator manufacturer (Maytag) to investigate using vacuum panel insulation in refrigeration units. Greenpeace has commissioned a participant of the GEA research team to determine the costs and benefits of this new technology applied to European style refrigerators. Computer simulations of the performance were conducted using the same software the GEA used for their own engineering analysis. The results show that the energy consumption of a typical 1992 model of refrigerator would be reduced by almost 6 times (-83%) using the Owens-Corning product (1" thick panelling) for the door and cabinet insulation (see Figure 1). This is achieved despite maintaining all the other design options constant, including the door seals. The projected cost of the Owens-Corning vacuum panel insulation would be a maximum of $165 U.S.-dollars for the whole refrigerator and the expected savings in electricity consumption for a twelve year lifespan are 3000 kWh which would cost a consumer 390ECU at current average EU electricity prices (there is near parity between the value of the ECU and the U.S. dollar). These figures show how even today a six fold improvement in energy efficiency can be attained from improving the insulation alone with a net cost benefit for the consumer. The performance of the vacuum panels will certainly improve from the values used in the simulation and the costs can be expected to fall substantially as the technology becomes more mature. There are a number of competing vacuum panel technologies which have substantially lower production costs but currently have worse insulation and durability properties than the Owens-Corning product, which could also offer an attractive solution. According to Tom Colley (1993), who has just completed an assessment of vacuum panel technologies for the U.S.DOE, vacuum panels with 70% of the energy performance of the Owens-Corning product, will cost 28% of the price. Greenpeace commissioned TNO to conduct a proper least life cycle cost analysis for a refrigerator assuming these type of vacuum panels to be commercially available. The results show that the least life cycle cost occurs for a design consuming only 17% of the energy of the base-case model and with a pay back period of only 1.3 years! The vacuum panel technologies also enjoy the advantage that they are more compact than conventional insulants: they maintain a much more stable temperature throughout the cabinet which enables much better food preservation properties, they reduce the build up of condensation (thereby saving on defrost time and the energy required to do it). Furthermore, in the future it is possible that consumers could buy the electricity for refrigerators at a reduced rate because vacuum panels would allow the use of off-peak electricity and the flattening of the domestic demand profile (this offers a double saving because the temperature of the kitchen will be cooler during the off-peak charging periods). IMPACT OF STANDARDS ON THE MARKET AND MANUFACTURERS: THE U.S. EXPERIENCE Despite some manufacturers opposing the implementation of standards the evidence of the National Appliance Energy Conservation Act (NAECA) in the U.S. is that imposing tough minimum energy efficiency standards does not adversely effect manufacturers or the market. In 1990 the US DOE served notice of standards to be implemented by 1993 which would reduce the average refrigeration unit consumption from between 25% and 30%. For all classes of refrigerators, fridge-freezers, and freezers, only 7 of the 1990 models out of 2114 met the 1993 standards. New standards for 1998 are due to be announced in 1995, but are expected to increase energy efficiency by an extra 25 to 50% (Remich 1993). The U.S. manufacturers complained bitterly about the presumed interference in the market and the huge cost of implementing the measures. Nonetheless, the standards have been met and exceeded with only a 1% increase in cost of the appliances to the consumer (Turiel 1993). The number of models available has not decreased, total sales have increased (+2.8% in Appliance (July 1993)) and the manufacturers appear to be performing at least as well as they were. After standards were implemented however the manufacturers have reacted in a positive fashion. Timothy Somheil, the senior editor of Appliance (the US manufacturers journal) wrote "The industry was well aware of the standard (the NAECA) and had no serious problems developing product to meet it" (Appliance: February 1993). Consumer acceptance of the new products is also reported to be high (Turiel 1993). Mr Lemser of Frigidaire states "...there will be an increasing distance between new products and the refrigerators of only 6 or 8 years ago. We will be able to demonstrate the savings vividly to the prospective buyer." (Appliance: February 1993). The Super Efficient Refrigerator Program (SERP), or "Golden Carrot", has demonstrated the potential for yet further energy savings in US domestic refrigerators. SERP's challenge to manufacturers was to bring into commercial production new models requiring 25% to 50% less power than current machines. At stake was a prize of $30 million in orders from electric utilities. The utilities' interest in encouraging customers to save energy was to avoid making the huge capital investments otherwise required in building new power plants. The super-efficient fridges incorporate the same features as energy-wasteful models of the same size. The only noticeable difference for consumers will be the reduced size of their electricity bills. The one notable deficiency of SERP was, while it excluded models using CFCs, it failed to bar use of other ozone-destroying chemicals such as HCFCs and environmentally-damaging chemicals such as HFCs. Many U.S. manufacturers now accept that the introduction of standards for both CFCs and energy has acted as a spur toward Accelerated Design Strategies which have reinvigorated their technical position relative to foreign producers. Timothy Somheil states "To many producers, energy efficiency and CFC-phaseout standards are seen as exciting challenges, and are being tackled aggressively" (Appliance: Feb 1993), he goes on to state "But in several years, when producers in Europe and other parts of the world could very well be striving to meet their own energy efficiency standards, they may find themselves turning to the U.S. for technology that its refrigerator/freezer industry is already being pushed to develop". There is growing evidence that U.S. manufacturers are now looking to penetrate lucrative overseas markets (especially eastern Europe) as they perceive that they have attained a new competitive edge over west European manufacturers. The U.S. experience demonstrates several points: ù Tough standards which have the effect of eliminating all existing models can be introduced successfully within a three year lead time. ù Manufacturers complaints should not be taken too seriously. ù Standards can often act as a stimulus for better designs. ù The CFC phase-out requirements are not contradictory to minimum energy efficiency standards. ù It is important to establish longer-term standards based upon a least life cycle cost engineering analysis (or similar) which are subject to a regular review. ù Energy efficiency procurement programmes such as SERP are of benefit and can be funded through electric utilities. HISTORICAL EFFICIENCY IMPROVEMENTS Since 1970 the historical improvement in energy efficiency of refrigeration units has been approximately 2.5% per annum. This has been attained predominantly through natural technical development as manufacturers seek to attain a competitive edge. Were this rate of efficiency improvement to continue (as argued below) then the average efficiency of new refrigeration units would have improved by almost 12% over the 1992 level by 1997. This outcome would render the -10% standard almost useless if implementation is delayed until 1997. The market factors influencing improvements of energy efficiency are complex and are not necessarily continuous. The GEA assumed (but did not justify) that energy efficiency would only increase at 1% per annum from 1992 to 2015. This value seems difficult to defend given the historical precedent. It has been argued that past energy efficiency improvements were largely a consequence of manufacturer perceptions of the oil price shocks of 1973 and the early 1980's. However, since the late 1970's oil has only had a minor role in European electricity generation and the price of electricity is relatively unaffected by the oil price. In 1980 German manufacturers entered into a voluntary agreement with their government to improve the energy efficiency of refrigerators and freezers by 15% to 20% over the period to 1985 compared to 1978. These targets were exceeded by a substantial margin. However over the same period in the U.S. average refrigerator efficiency improved by about 25% without national standards. It is important to remember that, although the European and U.S. markets are mainly internal, in that very few models are imported into either market, strong links exist between them, particularly through Electrolux and Whirlpool. Electrolux control over 20% of the European market and own the major U.S. manufacturer Fridgidaire, while Whirlpool have about 12% of the European market and are one of the top five U.S. manufacturers. Other big U.S. manufacturers also have stakes in European companies, e.g. General Electric Appliances who have 50% control of Hotpoint in the U.K. For these reasons one can expect a rapid technology transfer to occur between the U.S. and Europe, such that some of the energy efficiency solutions pioneered in response to legislation in the U.S. will be implemented in Europe even without mandatory standards. Aside from the proposed mandatory energy efficiency standards, the EU is likely to introduce mandatory energy and eco-labels for refrigeration units which should heighten an already growing awareness among consumers of the importance of energy efficiency. Fiscal incentives for energy efficiency can also be expected, for example through the EU's carbon/energy tax. If we consider these points collectively with the known technical potential for further refrigerator energy efficiency improvements, it seems difficult to justify a slow-down in the rate of annual efficiency increase that would occur without mandatory standards. THE GREENPEACE STANDARDS The evidence summarised above illustrates the technical and economic case for substantially tougher standards than the EU proposal. Following a careful consideration of the practical difficulties of introducing tough standards within a limited time frame, the following alternative standards are proposed (note that the percentage reductions cited below refer to the average energy consumption of the stock of new models against the average energy consumption of the stock of new models in 1992). ù A -15% standard to be announced by January 1st 1994 and introduced on January 1st 1995. ù -44% to -50% standards (see column 3, Table 7.1 in the technical and economic analysis) to be announced by January 1st 1994 and introduced for January 1st 1997. ù A review of standards to commence January 1st 1995 and to deliver advance notice of new standards by January 1st 1997 to be implemented between January 1st 2000 and 2002, depending on the manufacturer impact analysis. The case for introducing the -15% standard by 1995 was argued by the GEA. French manufacturers (whose models are of average efficiency) are reputed to have said that they could improve the efficiency of their models by 20% simply through using a better, currently available compressor (Lebot 1992). This remark indicates the ease with which those manufacturers whose models do not currently meet the -15% target could adapt their range within a short notification time and at low cost. The -44% to -50% standards are based upon the Least Life Cycle (LLC) cost analysis of the GEA. The intention is to set the standard line at a value which would result in the average new model having the same energy efficiency as is implied by the LLC cost. It is argued that manufacturers will on average produce models which better the standard line by 10%, which results in setting a standard line that is 10% weaker than the GEA long-term standard lines i.e. equivalent to multiplying the GEA long-term standard lines by about 1.1. The -44% to -50% standards are to implemented in 1997 rather than 1999, as was proposed by the GEA, partly because they are slightly weaker, but mainly because of the U.S. experience which showed that three years advance notice was adequate for manufacturers to produce a completely new range of appliances without great difficulty. The requirement for a review is important to ensure that standards keep pace with technological developments. Enough evidence has already been presented to suggest that the GEA least life cycle cost analysis is only adequate as an interim measure as vacuum panels were not considered. By 1997 vacuum panel technology should be sufficiently mature to set standards based upon its imposition for 2000. IMPACT OF STANDARDS ON COSTS AND EMISSIONS The impact of the efficiency standards was assessed using the energy savings forecasting model developed by the GEA. In addition the resulting pollutant emissions were also estimated using an adaptation of the GEA analysis. The Greenpeace emissions analysis differs from the GEA analysis in that it is assumed that refrigerators have a comparatively flat load profile and therefore only consume base-load electricity, which has implications for the fuel mix assumptions. Furthermore, the assumption of a 1% per annum efficiency improvement for the base-case (no standards) scenario is rejected in favour of a 2.5% per annum improvement. Otherwise the two methodologies are identical. The results of the Greenpeace A standards proposal for -44% to -50% standards in 1997 (giving a brand-type weighted average of -46% for the whole EU), Greenpeace B standards (the same as Greenpeace A up to 1997 but with a -80% standard in 2001), and the Commission's own current -10% proposal are shown for electricity consumption and CO2 emissions savings in Figures 2 and 3. The results are shown by comparison with the 2.5% per annum base-case efficiency improvement and also for the GEA base-case assumption of 1% per annum efficiency improvement. The two Greenpeace minimum energy efficiency scenarios are explained in detail in the main technical and economic analysis. The difference between the proposals is dramatic. The Greenpeace A standards could reasonably anticipate to be saving between 11 TWh and 15 TWh per year across the EU by 2000 growing to between 20 TWh and 34 TWh per year by 2010. The Greenpeace B standards will save between 14 TWh and 18 TWh per year across the EU by 2000 growing to between 44 TWh and 58 TWh per year by 2010. In contrast the Commissions standard would only introduce any benefit were the natural rate of efficiency improvement in the absence of standards to be less than 1.5%. Even if the pessimistic 1% per annum base-case assumption is accepted the results are very poor, saving only 2 TWh/year by 2000 rising to 5 TWh/year by 2010. In cost terms the Greenpeace A standard saves electricity of cumulative worth between 4.5 and 6.1 Billion ECU by the year 2000 and between 27.3 and 40.3 Billion ECU by 2010. The Greenpeace B standard saves electricity of cumulative worth between 5.1 and 6.7 Billion ECU by the year 2000 and between 46.8 and 59.8 Billion ECU by 2010. The Commission's proposal merely saves between 0 and 0.5 Billion ECU by 2000 and from 0 to 4.6 Billion ECU by 2010. The cumulative CO2 savings for the Greenpeace A standard are between 18 to 25 million tonnes of CO2 by 2000 and between 115 to 170 million tonnes by 2010. The cumulative CO2 savings for the Greenpeace B standard are between 21 to 28 million tonnes of CO2 by 2000 and between 199 to 254 million tonnes by 2010. For the Commissions proposal the cumulative savings are between 0 and 2 million tonnes of CO2 by 2000 and between 0 and 19 million tonnes of CO2 by 2010. ENFORCEMENT OF THE STANDARDS Proper enforcement of the standard is crucial if real energy savings are to be realised. Enforcement falls into two stages a) initial model certification; and b) inspection at point of retailing. At present, manufacturers produce their own energy consumption statistics measured according to the testing standard EN153. The EU proposal is that manufacturers should be allowed to test the performance of their own models for initial certification and that there should be no inspection at the point of retailing. Thus, the success of the standard would seem to be reliant upon rival manufacturers testing each other's models for compliance to ensure that the standards are not breached. The danger with this lowest-cost option is that testing may be ad hoc and that manufacturers may form alliances to cover-up non-compliance. Greenpeace have acquired independent test data from the U.K. Consumers Association who tested 383 models of fridge-freezers and freezers sold on the U.K., Dutch and Belgian markets. The results indicated that the manufacturers recorded energy consumptions which were on average 10.5% lower than the Consumers Association values. European manufacturers were invited to comment on the results at a meeting in October 1992 and claimed that the discrepancy was caused by a slightly different measurement method used in the U.K. test facilities (Peruzzo 1993). The reasoning presented has been strongly rejected by the testers at the U.K. Consumers Association (Larder 1993). This evidence was presented to the GEA and to the Commission, but does not appear to have been addressed in the draft Directive. It seems imperative that independent testing be conducted for both the certification and point-of-sale stages to ensure proper compliance. Greenpeace proposes the following: ù An independent agency should be established by the Commission to conduct both certification and random point-of-sale testing. ù Three models should be supplied by the manufacturer for initial certification with the energy rating being the average recorded under EN153. ù Should random point-of-sale testing indicate a model exceeding the standard by more than 10%, a further 10 models should be tested at random. If the average value of all ten models is found to exceed the standard by 5%, the model should be withdrawn from sale pending appeal. ù The agency should have the power to fine a manufacturer if the manufacturer is found to be guilty of deliberately trying to deceive the agency during the certification or point-of-sale testing programmes. CONCLUSIONS The Commissions draft Directive for minimum energy efficiency standards for refrigeration units has been shown to be ill founded and to have been contrary to the technical evidence which was presented to the Commission. Little or no economic and environmental benefits will occur through introduction of such a weak standard. Even under the most favourable assumptions the draft Directive would achieve very little toward the goal of stabilising EU CO2 emissions at 1990 levels by 2000. The Commissions failure to set tougher standards for the medium-term when all the necessary evidence was presented is of particular concern. The Commission's standards fall a long way below those proposed in the Dutch notification (-15% within 10 months of announcement). If the draft Directive were to be accepted the Dutch government would have been obliged to substitute the introduction of meaningful -15% standards which would have come into force in 1992 for a -10% standard whose implementation is delayed until 1997. This must raise the likelihood that the draft Directive will not be acceptable to the Dutch government. The standards which Greenpeace have proposed are wholly achievable, without deleterious consequences for the consumer or manufacturer, and would make a significant difference to domestic energy consumption in a relatively short time. TECHNICAL AND ECONOMIC ANALYSIS 1: MANUFACTURERS AND MARKETS 1.1 THE EUROPEAN MARKET Of the three principal markets for electrical appliances in the world, Europe (EU & EFTA) is the largest, worth around 25 billion ECU annually (with another 7 billion ECU in East Europe) (Euromonitor 1993). This compares with the US and Japanese markets worth 15 and 12 billion ECU respectively. For refrigeration units the three major regional markets have quite different characteristics and are predominantly supplied by local manufacturers. European appliance manufacturers can be divided between those who specialise in small appliances and those who specialise in large appliances including fridges and freezers. Often appliance manufacturing is only one component of larger corporate operations, i.e. some appliance operations belong to groups specialising in electronics (e.g. Philips (now Whirlpool), Thomson and General Domestic Appliances (GDA)), while others may belong to engineering firms (e.g. AEG, Bosch-Siemens, De Detrich, Fagor and Liebherr). The leading manufacturers of large appliances also dominate the market of household fridges and freezers, although there are many niche markets for smaller manufacturers. Unlike the concentrated markets of the US and Japan, the European domestic refrigeration market remains relatively fragmented with well over a hundred brands and about 40 independent manufacturers. The market leaders Electrolux, Bosch-Siemens and Whirlpool, account for less than half of the large appliance market. In comparison, in the US the top five manufacturers (Whirlpool, General Electric, Electrolux, Maytag and Raytheon) account for 97%, by volume, of the total large appliance market. The Japanese market is also highly concentrated, being dominated by just three manufacturers (Matsushita, Hitachi and Toshiba). The fragmented nature of the European market is a result of the multi-national character of Europe, wherein historically most European countries have evolved at least one white-goods manufacturer who has generally tailored product to the specific needs and conditions of the home market. However, the advent of economic harmonisation within Europe, through the EU and EFTA trade groups, has resulted in increasing product standardisation and the opening of national markets within the EU. This is exposing the smaller national based manufacturers to greater competition from the multi-national companies. Consequently there have been a number of important takeovers over the last 10 years as the European market becomes more concentrated. Some significant takeovers have been from US manufacturers seeking to take advantage of the new opportunities that the changing trading conditions present. 1.1.1 PRODUCT CATEGORIES The two general classes of domestic food refrigeration technology currently on sale in the European market are compressor-type refrigerators and absorption-type refrigerators. Compressor-type refrigeration is utilised in about 95% of appliances while absorption-type refrigeration accounts for the remainder, mainly for markets where low noise is a premium, e.g. hotel mini-bars and caravans. This report is only concerned with the more common-place compressor-type refrigeration appliances. There are three general categories of these appliances; refrigerators, freezers, and combinations known as fridge-freezers. These categories are sufficient for discussing markets but are too aggregated to use as a basis for product grouping to be applied for efficiency standards. This aspect is discussed in detail in Section 3.2. In some respects European refrigeration appliances are distinct from those available on the North American and Japanese markets. European and Japanese appliances tend to be smaller than American models, partly as a consequence of the smaller kitchen space available. While Japanese fridges often have a greater number of compartments with different operating temperatures than European fridges, reflecting differing consumer preferences toward food storage and preparation. 1.1.2 NATIONAL MARKETS According to 1991 sales data supplied by the Italian manufacturers association (ANIE 1993) and shown in Table 1.1 the largest market for all refrigeration appliances was Germany with 4.6 million units. France, UK, Italy and Spain also feature strongly. These five countries account for over 80% of the total EU sales. Table 1.1: Sales of Refrigeration Appliances in the EU for 1989 and 1991 *1. Nation Model Sales (thousands of units) Fridges Freezers Fridge/Freezer Total 1989 1991 1989 1991 1989 1991 1989 1991 Be 217 142 160 114 136 85 513 341 De 1735 3050 750 1330 345 272 2830 4652 Dk 129 117 87 96 58 77 274 290 Es 1836 686 330 229 - 545 2166 1462 Fr 1551 845 714 750 450 1166 2715 2761 Gr 350 350 - 374 - 340 350 1064 Ir 40 40 56 56 19 19 115 115 It 830 980 350 440 680 800 1860 2220 Lu 10 10 12 12 10 10 32 32 Nl 360 290 200 180 118 205 678 675 Po 245 78 115 114 - 263 360 455 UK 780 1084 584 810 836 823 2200 2717 * 1 The 1989 data is taken from the GEA report (GEA 1993), while the 1991 data is taken from the Italian Ecolabel study (ENEA 1993). 1.1.3 SALES TRENDS The trend in EU sales between 1985 and 1991 is shown in Table 1.2. The 1991 data is from a different source to the 1985-1989 data and shows a significant difference between the 1989 data for the Freezer and Fridge-Freezer data. It is not clear whether this is caused by genuine changes in sales, or by German unification and the inclusion of East German sales data, or by inconsistencies in reporting sales between the two sources. The general picture which emerges from the internally consistent 1985 to 1989 data is that sales within the EU are only growing very slowly and that the market is mature. Table 1.2: Sales of Refrigeration Appliances in the EU between 1985 to 1991 *1. Appliance EU Model Sales (thousands of units) Category 1985 1986 1987 1988 1989 1991 Fridge 6950 7292 7623 7913 8083 7672 Freezer 2995 3172 3262 3326 3358 4605 Fridge- Freezer 2379 2465 2602 2638 2652 4505 Total 12324 12929 13487 13877 14093 16784 *1 The 1989 data is taken from the GEA report (GEA 1993), while the 1991 data is taken from the Italian Ecolabel study (ENEA 1993). 1.1.4 OWNERSHIP AND MARKET MATURITY Saturation data showing the percentage of households owning refrigeration appliances by country are given in Table 1.3. The high saturation levels indicate the maturity of the market and the limited prospects for sales volume growth, because sales are dominated by model replacements. However the figures do indicate some significant differences in consumer preferences by nation e.g. the comparatively low ownership of fridge-freezers and freezers in Portugal and Greece, or the stronger preference for fridge-freezers in Italy and Spain than Germany or Denmark. There are often significant differences in household appliance ownership within national boundaries too, e.g. a preference for chest freezers in the south of Germany compared to the north (EIU 1990), and a preponderance of fridges in the former east Germany, or the far lower ownership of freezers in Paris than the other French regions (EIU 1991). Table 1.3: Household Saturation Levels for Refrigeration Appliances Appliance Household Saturation Levels (%) by Country Be De Dk Es Fr Gr Ir It Lu Nl Po UK Year 1990 1990 1990 1990 1990 1990 1990 1990 1990 1991 1990 1990 Frid ges 54 73 62 43 58 70 54 42 55 60 65 48 Free zers 62 66 63 30 43 27 58 31 62 46 30 39 Fridge- Free zers 53 36 38 51 40 24 43 62 54 45 29 53 1.2 PRODUCTION WITHIN THE EU 1.2.1 LEADING EU MANUFACTURERS Table 1.4 shows the leading European fridge and freezer manufacturers, their country of origin, market share and main brands. Electrolux of Sweden, with its large number of national brands, has the largest share and accounts for almost 1 in 5 of all model sales in Europe, although the majority of these are manufactured outside Sweden. Whirlpool International and Bosch-Siemens of Germany are roughly joint second. The American giant, Whirlpool, became a major player in the market in 1984 through the acquisition of Philip's Major Appliance Division. According to market share data from the Italian manufacturers association (ANIE 1993), shown in Table 1.4, the top three refrigerator manufacturers (Electrolux, Bosch-Siemens and Whirlpool) have 42.2% of the total EU market while the next seven leading manufacturers have a 26.8% share, the remaining 31% is shared between about 30 other manufacturers. 1.2.2 BRAND NAMES Most of the top manufacturers sell their products under a variety of brand names. Some of these are the names of genuine manufacturers who have been acquired by the larger corporation, but often a one-time traditional national brand name is used to capitalise on a historical resonance that is perceived to exist within a specific market. Brand names are also useful for manufacturers to encourage consumers to differentiate between their product range, e.g. different brands might be used for up-market and down-market models. The top eleven manufacturers own 88 brands, listed in Table 1.4, many of which are household names and leading sellers throughout Europe. Electrolux alone owns at least 30 different brands. The huge variety of brand names sold within Europe complicates the analysis of the European market and the comprehension of manufacturer and governmental attitudes toward standards. Table 1.4: Market Shares, Subsidiary Companies and Brands of the Leading European Refrigerator Manufacturers *1 Company Country EU Subsidiary Companies & Brands of Market Origin Share % Electro Sweden 16.6 Acec, Arthur Martin, Atlas, lux Buderus,Castor, Corbero, Cylinda, Edesa, Electrolux, Elektra, Elektro Helios, Faure, Frigidaire, Husqvarna, Juno,Marynen, Newpol, Nordton, Rex, Rosenlew, Scandinova, Tappan, Therma, Tricity-Bendix, Vestfrost, Voss, White-Westinghouse, Zanker, Zanussi, Zoppas, Minority holding in AEG Bosch- Germany 13.6 Agni, Asko Polar, Atag, Siemens Balay, Bosch, Constructa, Corcho, Crolls, Linx, Neff, Pelgrim, Pitsos, Safel, Siemens, Super Ser Whirlpool US 12.0 Bauknecht, Erres, Ignis, Laden, Philips-Whirlpool, Phonola, Radiola, Ruton, Sierra/Laden, Whirlpool Merloni Italy 5.5 Ariston, Blue Air, Colston, Indesit, Hirundo, Philco, Scholtes, Smeg AEG Germany 4.4 AEG Thomson France 3.8 Brandt, De Dietrich (49% share), Sauter, Thermor, Thomson, Vedette GDA UK 3.2 Hotpoint, Creda, Jackson Candy Italy 2.8 Candy, Gasfire, Kelvinator, Rosieres, Zero Watt Fagor Spain 2.1 Aspes, Fagor LEC UK 2.0 LEC Miele Germany 1.5 Miele Ocean Italy 1.5 Alaska, Atlantic, Ocean, Samet, San Giorgio, Westen, Westinghouse *1 Market share data from (ANIE 1993). Other independent manufacturers include: Liebherr, Gaggenau, Blomberg, Foron (Germany); Gorenje (ex-Yugoslavia); De Dietrich (France); Gram, Derby (Denmark); NEI (Hungary); Maytag (USA/UK); Norfrost (UK); Snowcap, Minsk (Russia). In addition to the above listed makes and brands there are a number of brands owned by distributers under-which a variety of manufacturers models are sold. 1.2.3 RECENT MERGERS AND TAKEOVERS IN EUROPE The number of independent manufacturers within Europe has steadily declined in the last decade as the larger manufacturers have sought to consolidate their grip on the market through a number of important acquisitions, listed in Table 1.5. Table 1.5: Major Acquisitions in Europe 1984-19921 Purchasing Company Country of Company Acquired Country of Origin Origin (1984) Electrolux Sweden Zanussi Italy (1985) Electrolux Sweden Zanker Germany (1987) Candy Italy Rosieres France Electrolux Sweden Thorn EMI Large UK Appliance Division GEC UK Creda UK (1988) Electrolux Sweden Corbero/Domar Spain Merloni Italy Indesit Italy Whirlpool US Philips Major Netherlands Appliance Division 2 (1989) Bosch-Siemens Germany Balay/Safel Spain Electrolux Sweden Buderus Household Germany Appliance Division Merloni Italy Scholtes France Maytag US Hoover UK (1991) Electrolux Sweden Lehel Hungary Candy Italy Otsein Spain 1 Source (Euromonitor 1993) 2 Deal completed 1991 The medium-sized manufacturers have reacted to these takeovers by seeking strategic alliances to enable them to move beyond their domestic markets. Some of these moves, from Euromonitor (1993), are described below. ù In 1989 General Electric of the US took a 50% share in the household appliances division of GEC (UK), to become General Domestic Appliances (GDA). ù In 1991 Thomson Electromenager took a 50% stake in the electrical appliances division of De Dietrich, creating De Dietrich Europeenne d'Electromenager. ù In 1991 Thomson formed an alliance with Fagor of Spain. the deal does not involve cross-shareholdings but provides for co-operation in components sourcing, product exchange, and R&D. ù In 1992 GDA joined the Fagor-Thomson-De Dietrich grouping (again no cross-shareholdings were involved). ù In 1992 Electrolux took a 10% stake in AEG Hausgerate. The companies are to continue to compete in the German market but co-operation in production and development is envisaged. 1.2.4 MANUFACTURER MARKET SHARES BY NATION The distribution of manufacturer market shares for the largest national markets are given in Table 1.6. It can be seen from this table that the national markets have quite different character. The German and Italian markets are dominated by indigenous manufacturers. In France, Spain and the UK indigenous manufacturers are also market leaders, but there is far greater competition from the rest of Europe. Furthermore, the French, Spanish and UK manufacturers only sell strongly in their own national markets while the Italian and German manufacturers are strong exporters. Thus, the largest French and UK manufacturers, Thomson Electromenger and General Domestic Appliances (GDA), have only have 3.8% and 3.2% of the EU market respectively. In EU nations without an indigenous industry, such as the Benelux countries, the large multinational corporations, e.g. Electrolux and Whirlpool, feature very strongly. Table 1.6: Market Shares for the Largest National Markets1 FRIDGES AND FRIDGE-FREEZERS Germany 1989 France 1990 UK 1991 Quelle 19% Thomson 25% Hotpoint 23% Bosch 15% Whirlpool15% Lec 19% Siemens 13% Electrolux 14% Zanussi 25% AEG 13% Sidex - Own-Brands 10% Liebherr11% Ariston - Electrolux 9% Bauknecht5% Candy 7% Philips 5% Whirlpool 4% Merloni 2% Italy 1990 Spain 1988 Whirlpool 24% FagorGroup 34% Zanussi 23% Bosch- 22% Candy 19% Siemens 15% Ariston 14% Electrolux 14% Indesit 10% Whirlpool 1 Sources (GEA 1993; EIU 1989; G&A 1992) FREEZERS Germany 1989 France 1990 UK 1991 Liebherr 22% Thomson 23% Hotpoint 14% Quelle 18% Whirlpool 15% Electrolux 12% Bosch 15% Electrolux 10% Own-Brands 12% Siemens 12% Bosch- Lec 9% AEG 11% Siemens 5% Zanussi 9% Bauknecht 5% Ariston 2% Norfrost 9% Bosch- Siemens 6% Whirlpool 6% Italy 1990 Spain 1988 Candy 26% Electrolux 43% Zanussi 25% Whirlpool 16% Smeg 15% Fagor Group 10% Merloni 14% Bosch- 5% Indesit 8% Siemens 1 Sources (GEA 1993; EIU 1989; G&A 1992) 1.2.5 NATIONAL PRODUCTION, EXPORTS AND IMPORTS By far the largest producing countries are Italy and Germany, producing about 75% of the fridges and fridge-freezers sold in the EU and about 55% of the freezers. Italy alone produces about 44% of all models sold within the EU while Germany produces 29%. Table 1.7 shows the production, exports and imports by producing country in the EU. The final column gives the apparent market, the production plus imports minus exports, which corresponds reasonably well to the actual sales data in Table 1.1. The same data is shown graphically in Figures 1.1 & 1.2. Table 1.7: Model Production, Exports and Imports by Producer Neighbour *1 Prod Exports Imports Apparent Exports/ Imports/ Exports/ uction Market Production APPM Imports Fridges & Fridge- Freezers Denmark 257 161 140 236 63% 59% 1.15 (1989) 451 150 1695 1988 33% 85% 0.09 France(1989) 3374 1583 1376 3167 47% 43% 1.15 Germany 4695 2985 255 1965 64% 13% 11.71 (1990) 335 160 100 275 48% 36% 1.60 Italy(1989)852 124 137 865 15% 16% 0.91 Portugal 1200 150 852 1802 13% 47% 0.18 (1991) Spain (1988) U.K. (1991) Production Exports Imports Apparent Exports/ Imports/ Exports/ Market Product APPM Imports Freezers ion Denmark 777 669 17 125 86% 14% 39.35 (1989) 177 74 628 731 42% 86% 0.12 France(`89)663 377 828 1114 57% 74% 0.46 Germany 1460 1052 89 497 72% 18% 11.82 (1990) 200 110 20 110 55% 18% 5.50 Italy(`89)243 153 54 144 63% 38% 2.83 Portugal 429 107 389 711 25% 55% 0.28 (1991) Spain (1988) U.K. (1991) *1 Sources (GEA 1993; EIU 1989) It is clear from this data that the most self-reliant countries (those with the lowest percentage of imports as a proportion of their domestic market) are Italy, Portugal, Germany and Denmark and Spain. The least self-reliant are obviously those with no domestic producers: The Netherlands, Belgium, The Republic of Ireland, Greece and Luxembourg. However, France and to a lesser extent the UK import a large proportion of the models they consume. The greatest exporter nations, Italy and Germany, export 4037 and 1960 thousand units per annum respectively, but Denmark also exports 860 thousand units. Spain, Portugal, UK and France are substantially less important, exporting about 250 thousand units each per annum. 2: ENERGY MEASUREMENTS AND DATA 2.1 MEASURING REFRIGERATOR ENERGY CONSUMPTION For policy decisions to be taken concerning refrigerator energy consumption it is important that an acceptable method of measuring refrigerator energy consumption be defined. This measurement method should be straightforward to apply, easily replicable and representative of the actual usage conditions for the appliance. Satisfaction of all these criteria is difficult to achieve. The most universally accepted appliance measurement methods are defined by the International Standards Organisation (ISO). The standards they produce relating to refrigerator product classes, storage temperatures, storage volume and energy consumption are: ù ISO 7371-1987: Household refrigerators with and without frozen food compartments, ù ISO/DIS 8187.3 - 1988: Household refrigerator- freezers, ù ISO 5155-1983: Household freezers. In Europe most countries and their manufacturers report refrigerator energy consumption according to the European standard EN153 (1991), "Methods of measuring the energy consumption of electric mains operated household refrigerators, refrigerator-freezers, frozen food storage cabinets, food freezers and their combinations, together with associated characteristics". EN153 is issued by the European standards organisations CEN and CENELEC, but it is based upon the ISO standards. There is an ongoing process of standards harmonisation taking place within Europe, such that most EU and EFTA members have either adopted EN153 are about to do so. Of those countries which have not yet done so: France, UK, Belgium, Ireland and Luxembourg; France and the UK have energy standards that are effectively identical to EN153. According to EN153 the standard energy consumption is the amount of electricity a refrigerator consumes over a 24 hour period when placed in a room with a temperature of 25oC and a relative humidity of 45% to 75%. The fresh-food compartment is empty during the test while the frozen-food compartment is filled with test packages. The fridge temperature must average 5oC ñ0.5oC and must always be in the range 0oC to 10oC. The temperature of the frozen food compartment or freezer compartment must not exceed the values given in Tables 3.2 and 3.3. The doors to all compartments are closed throughout the test. EN153 also defines the means of measuring the volume of refrigerators that distinguishes between the gross volume, defined as "the total volume within the exterior walls of the appliance" and the net volume, defined as "that part of the gross volume that remains after subtraction of components and spaces not suitable for food storage". It follows that energy consumption should be related to the net volume as this connects the amount of energy consumed to perform a unit food storage task. EN153 is an imperfect measure as it only partially imitates real refrigerator usage patterns. It does not take full account of probable food loading and unloading conditions, and the choice of 25oC as the ambient operating temperature may not be typical of true usage circumstances. However, these deficiencies can not be rectified until better in-situ usage data is available. At the present time there is uncertainty as to whether EN153 under-estimates or over-estimates the energy consumption of a refrigerator compared to it's typical in-situ usage in European households. Better quality information will only become available subsequent to a thorough, European wide, in-line monitoring survey. Despite these short-comings it is appropriate to use EN153 as the energy measurement basis for European minimum energy efficiency regulations because it is the most widely accepted reproducible measurement standard and is the easiest to administer. Although EN153 may not record the energy consumption of a refrigerator as it may be experienced in practice, it is likely to give results proportional to the in-situ value and therefore is satisfactory for the comparative purposes involved in regulation. Unless otherwise stated the refrigerator energy consumptions and net volumes referred to in this report are those measured using EN153. 2.2 ENERGY DATA FROM THE EUROPEAN REFRIGERATION MARKET This study makes use of two main sources of refrigeration appliance energy data: ù the GEA database, ù the UK Consumers Association database. Both databases give: the model make and mark, the national market where it is sold, the energy consumption and the net volumes of each compartment. The GEA database comprised 3761 models from the European market which is estimated to be about 75% of the models on the market in 1992. The number of models are distributed by nation, type and year as shown in Table 2.1. TABLE 2.1: THE DISTRIBUTION OF MODELS IN THE GEA DATABASE Country Fridges Freezers Fridge- Total Year of of sale Freezers data Germany 370 372 217 959 1990 Denmark 135 159 108 402 1991 Spain 32 49 92 173 1992 France 153 338 164 655 1990 Italy 78 126 248 452 1991/92 Nethe'ds 354 390 252 996 1990 Portugal 10 32 20 62 1990 UK 20 26 26 72 1992 All 1152 1492 1127 3771 1990-2 The GEA database is comprised of manufacturers measurements conducted according to the European test standard, EN153. The data from the UK is actually a subset of the Consumers Association database, but uses only the manufacturers measurements not the Consumer Association measured values. The UK Consumers Association performs independent testing of refrigerators for models sold on the UK, Belgian and Dutch markets. The number of models are distributed by nation, type and year as shown in Table 2.2. TABLE 2.2: THE DISTRIBUTION OF MODELS IN THE CONSUMERS ASSOCIATION DATABASE Country Fridges Freezers Fridge- Total Year of sale Freezers of data Belgium 7 15 18 40 1991/92 Nethe'ds 56 90 56 202 1990/92 UK 117 89 99 305 1991/92 All 180 194 173 547 1990/92 2.3 MANUFACTURER REPORTING OF ENERGY CONSUMPTION The issue of whether refrigerator manufacturers are properly recording their models energy consumption is crucial to the debate concerning the impact of minimum energy efficiency standards and of the method of policing standards. In order to evaluate the integrity of manufacturers measurements of their refrigerators energy consumption Greenpeace purchased an independently measured database of refrigerator energy consumption from the UK Consumers Association (described in Section 2.2). The database contains manufacturers measurements and Consumers Association measurements for 547 refrigerator models sold on the UK, Belgian and Dutch markets, between 1991 and 1992. Figures 2.1 to 2.6 show the energy consumption as a function of the adjusted volume (defined in Section 3.1) for refrigerator categories R2, R3, R4, R5 & R6, F1 and F2. For each refrigerator model the energy consumptions measured by the manufacturers and the Consumers Association are shown. Straight lines are fitted to each data set. It can be seen that for all categories but R5 & R6 the Consumers Association measure substantially greater energy consumption for the same models than the manufacturers. The average energy consumption value in each category is given in Table 2.3. TABLE 2.3: AVERAGE ENERGY CONSUMPTION BY REFRIGERATOR CATEGORY AS MEASURED BY THE CONSUMERS ASSOCIATION AND BY MANUFACTURERS Average Energy Consumption Model kWh/year Number of Category models *1 C.A. Manufacturers F1 520.61 467.13 49 F2 498.78 465.19 49 R4 613.79 545.43 81 R2 335.42 310.44 19 R3 346.01 331.18 30 R5 & R6 283.70 289.14 59 * 1 The number of models is less than the number in the full Consumers Association database because not all models had manufacturers measurements quoted. The largest disparity occurs for the fridge-freezers and freezer categories where the Consumers Association energy consumption measurements are on average 10.9% higher than the manufacturers measurements. Even with the R2 and R3 categories the average energy consumption measured by the Consumers Association is 5.8% higher than the manufacturers measurements. Only with the R5 & R6 category are the measurements approximately the same, with the manufacturer's measurements being on average 1.9% higher than the Consumers Association's. The fact that the disparity occurs only for models with frozen food compartments may be significant. The evidence suggests that where the energy consumption recorded is strongly dependent on the strict adherence to the compartment temperature constraints defined under the EN153 and ISO standards that the manufacturers underestimate the energy consumption of their models. This may well arise because they apply the temperature constraints less strictly than the standards demand. This evidence of mis-measurement was presented to representatives of the Italian manufacturers association, ANIE, in December 1992. A representative of Zanussi (Peruzzo 1993) claimed that the disparity was caused by different measurement practices followed by the UK Consumers Association, in particular: ù the usage of a slightly different mains voltage, ù the manufacturers practice of interpolating energy consumption depending on the temperature measured in the temperature controlled compartments. However, Duncan Larder, a test engineer at the Consumers Association examined these claims and concluded that the interpolation practice should not effect the results while the slight difference in mains voltage could only account for a maximum 2% difference. These findings suggest that manufacturers should not be relied upon to test their own models or to police each other, as has been proposed under the draft EU Directive. 3: MEASURES OF APPLIANCE ENERGY EFFICIENCY 3.1 ENERGY EFFICIENCY INDEX It is not sufficient to judge the energy performance of an appliance simply in terms of the total amount of energy it consumes, rather it is necessary to take account of the functionality of the appliance as well. This consideration necessitates the development of an energy efficiency measure which can be used as a basis for ranking appliances and providing a reference point for establishing standards. The measure should be as independent as possible of the basic function and utility of the appliance, in other words, the purpose of the measure is to factor out effects which are fundamentally associated with the utility of the appliance from the efficiency evaluation. For refrigeration appliances this means taking account of the food storage capacity as well as the intended operating temperatures of the storage compartments. Other important factors are the external dimensions of the appliance, whether the appliance is intended to freeze food, how the device is to be defrosted, and the climate within which it will be operated. Traditionally, analysts have compared the energy efficiency of similar refrigeration appliances through the energy consumed per litre of storage space. This treatment produces a distorted measure for two main reasons: ù it does not take account of differing compartment operating temperatures, ù it favours larger appliances over smaller ones. The first point arises because the cooling-load is approximately proportional to the temperature difference between the inside and outside of each appliance compartment and therefore it is not sensible to compare energy consumptions per unit storage-volume (the "specific energy" consumption) without making adjustments for differences in this value between compartments. The GEA (1993) and others have considered this issue and proposed using an "adjusted volume" which is defined such that the volume of the individual compartments are multiplied according to their operating temperatures and summed as follows: êi = (Tamb-Ti/Tamb-5) [3.1] n Vadj = äViêi i=1 where Vadj is the adjusted volume of the whole appliance, Vi is a given compartment's storage volume, Tamb is the ambient temperature during the energy consumption test (for EN153 this is 25oC to represent a typical kitchen temperature), and Ti is the given compartment's design operating temperature (Celsius scale), Wi is the volume adjustment factor for the given compartment, and n is the number of compartments. The value of 5 is used in the denominator because this is the usual operating temperature (in degrees Celsius) for fridge compartments where food is to be kept chilled, but not frozen. The adjusted volume for the whole appliance is the sum of the compartment adjusted volumes. Even this more sophisticated concept of adjusted volume implies a significant simplification of the true heat transfer processes, because it assumes that the myriad sources of heat flow between the compartments, the refrigerator and the ambient environment can be represented by a 1-dimensional linear heat-transfer process between each compartment and ambient. In particular, it assumes that each compartment is wholly surrounded by space at a single (ambient) temperature when in practice at least one side is likely to be connected to another compartment of the appliance. This mis-representation is often compounded because the insulation used in walls adjoining ambient is usually thicker than the insulation used for internal partitions between compartments. The adjusted volume also fails to properly account for edge effects, compartmental geometry, thermal-bridges, or other sources of non-linear heat-transfer, all of which will complicate an analysis. However, the adjusted volume concept is still of practical value because it offers a reasonable compromise between a thorough treatment of the actual heat transfer processes and the regulatory need for simplicity. Provided adjusted volume is only used to compare the specific energy consumption for products of the same basic type, once the impact of appliance size has been factored out (see the next paragraph), then it is an acceptable simplification. The second point arises because the cooling-load of a refrigeration appliance is likely to be proportional to it's surface area. Surface area is not proportional to the storage capacity, thus a smaller appliance has a larger surface to storage-volume ratio than a larger appliance. This means that, even though a smaller appliance may use less total energy than a larger appliance, if the insulation quality is the same between the two models then the smaller model will have a greater cooling-load per unit storage-volume and hence probably consume more energy on a unit volume basis. The practical importance of this phenomenon is illustrated through Figure 3.1 which shows the annual energy consumption of 1992 models of chest freezers against their adjusted volume. It is apparent from this figure that models with larger adjusted volumes consume more energy on average than smaller models. However, if the same data is plotted as the specific energy (annual energy consumption per unit adjusted volume) against the adjusted volume, see Figure 3.2, it is clear that the smaller models have, on average, higher specific energy consumption. As we wish to develop a energy efficiency measure which does not discriminate against a refrigerator on the basis of it's storage-volume it is necessary to factor out the effect of volume from the measure. A proper deterministic treatment of this phenomenon based on applying known physical principles would have to be done on a model by model basis. This would involve considerable complexity rendering it impractical for regulatory purposes. As a consequence analysts (GEA 1993; Turiel et al 1992, CA 1992) have favoured a less pure, but more applicable empirical treatment, called the "statistical approach". It is clear from Figure 3.1 that the dependency of energy consumption on adjusted volume is approximately linear for the given category of refrigeration appliance. Thus, it is possible to describe the average dependency for the product category by fitting a straight line to the data, known as the "reference line". In actuality the best fit is likely to be given by a shallow parabola, but the added complexity of using this type of fit scarcely seems justified for the slight improvement in accuracy gained. Once the reference-line equation of the average energy consumption for a given adjusted volume is known for a particular product category the energy efficiency can be computed. This is done by dividing the actual model energy consumption by the average energy consumption for its adjusted volume and product category. By way of example, consider a chest freezer with a net volume of 296 litres and annual energy consumption of 462 kWh/year. The operating temperature of the freezer compartment is -18oC thus from equation [3.1] the adjusted volume is 636.4 litres. The reference line describing the average energy consumption of chest freezers as a function of adjusted volume (see Table 5.1) is: EAV = 0.446 Vadj + 181 [3.2] where EAV is the average annual energy consumption (kWh/year). Thus, for a 296 litre chest-freezer the average annual energy consumption is 464.8 kWh/year. The energy efficiency index, EEFF is then given by: EEFF = EME/EAV [3.3] where EME is the measured energy consumption (kWh/year). From this equation, the energy efficiency index of the example freezer is 0.994, i.e. the freezer uses 99.4% of the energy of an average model in its class and adjusted volume. 3.2 PRODUCT CATEGORIES In order to apply the energy efficiency-measure methodology described in Section 3.1 it is necessary to define the categories of refrigeration appliances which can be considered to belong to the same product group with the same energy reference line. This is important so the energy efficiency index can be applied without discriminating against a refrigeration appliance for reasons concerned with its utility. The GEA categorised the refrigeration appliances as follows: TABLE 3.1. THE GEA REFRIGERATION UNIT CATEGORIES FOR ENERGY EFFICIENCY STANDARDS Code Category Vadj (FFC = Frozen Food Compartment) = Adjusted Volume in litres R1 Refrigerator with 1 Star FFC VRef+1.55*VFFC R2 Refrigerator with 2 Star FFC VRef+1.85*VFFC R3 Refrigerator with 3 Star FFC VRef+2.15*VFFC R4 Fridge-Freezer (Refrigerator with VRef+2.15*VFFC 4 Star FFC) R5 Refrigerator with Chiller VRef+0.75*VChil Compartment R6 Other Refrigerators VRef+1.25VFFC F1 Chest Freezer 2.15*VFre F2 Upright Freezer 2.15*VFre where VRef, VFFC, VFre and VChil are the net volumes of the fridge, the frozen food, the freezer and the chiller compartments respectively (in litres). The star rating system is used by manufacturers to grade the temperature and freezing rate of the individual food compartments, as defined in Table 3.2: TABLE 3.2: DEFINITIONS FOR THE STAR RATING SYSTEM Type of Temperature Performance characteristics compartment reached 1 star -6oC No freezing capacity (*) 2 star -12oC No freezing capacity (**) 3 star -18oC No freezing capacity (***) 4 star -18oC Freezing capacity above (****) 4.5kg/24hours per 100 litres net volume; with a minimum of 2kg. In addition to these type of compartments, which are intended to store frozen food and/or freeze fresh food, other higher temperature compartment types are also in use, Table 3.3. TABLE 3.3: HIGHER TEMPERATURE REFRIGERATOR COMPARTMENTS Type of Temperature Performance characteristics compartment reached 5oC Storage of fresh food of normal refrigera 0oC perishability tor Storage of fresh food of high 0oC*1 10oC perishability chiller*1 -6oC < T < Storage of fresh food with low 0 star*2 0oC perishability No freezing capacity or frozen food storage *1 There is some confusion concerning the definition of compartment categories used in the draft refrigerator eco- labelling study (ENEA 1993) and sponsored by the Italian manufacturers association. ENEA define a "chiller" compartment to be a 0oC compartment and define a compartment at 10oC to be a "cellar". These definitions differ from those accepted by the International Standards Organisation (ISO) who have no "cellar" compartment category while their definition of a "chiller" is cited in Table 3.3. *2 The International Standards Organisation (ISO) do not recognise this as a separate category, but the eco-labelling study assert that a number of appliances have compartments with these characteristics. The categorisation defined in Table 3.1 was arrived at following a series of statistical tests on the full refrigeration product database to assess whether it was valid to merge any of the categories. The database was comprised of manufacturers energy consumption data on 3761 models which is equivalent to 75% of the refrigeration models available on the European market in 1992 (see Section 2.2). With the exception of the R5 and R6 categories, the results showed that models in each of the defined categories exhibited significantly different adjusted volume or energy consumption characteristics from models in other categories and therefore no two categories could be reasonably combined into one. In addition to the categories of Table 3.1 the GEA also identified a significant difference in energy use for "no-frost" appliances. These models are designed to prevent frost build-up around the evaporator plate through the use of forced convection and/or periodic heating. A fan blows air over the evaporator and into the fresh-food compartment which reduces thermal stratification and lowers condensation around the evaporator plate. The evaporator often has an electric resistance heater which is operated periodically to melt the frost. Both the fan and the heater require extra electricity to operate and cause the cooling load to increase, such that on average "no-frost" models consume more energy than conventional models. The draft refrigerator eco-labelling study (ENEA 1993) disagreed with the GEA partition of the product categories in some areas. According to the draft eco-labelling study the GEA database and procedure can be questioned on the following grounds: a) the "cellar"1 reference temperature assumed was 10oC, while the ISO 8187 (ISO 1988) test standard assumes 12oC, this has the effect of altering W from 0.75 to 0.65 (equation [3.1]), b) no-frost appliances were not always excluded from the R4, F1 & F2 categories in the database because of a failure to identify them, c) appliances with no-frost compartments are considered as if the whole appliance is a no-frost appliance, while in actuality some models are a combination of normal and no-frost compartments, d) R5 (fridge with "cellar"1 compartment) and R6 (fridge with 0 star compartment) have been merged because appliances could not be discriminated in most of the GEA data collected, e) the database contains models from 1990, 1991 and 1992, 1 See subscripted comments under Table 3.3. The eco-labelling study appears to be referring to the GEA definition of a "chiller" compartment and not a "cellar". Points c) and e) appear to be valid observations and are discussed in Sections 5.3.2 and 9.2. The question concerning the proper chiller temperature for calculation of W (point a) is also justified and may have arisen because of some ambiguity in the ISO standard 8187. In Table 1 of the standard the chiller temperature is defined as 8oC to 14oC, but in Section 15.2.1 which discusses energy measurement procedure the value of 12oC is used. However, the value of W = 0.75 was used in formulating the reference line of Section 5.1 and it would introduce confusion to make modifications at this stage. Given that the chiller compartment volume is only a small proportion of the total appliance volume it is unlikely that this oversight will significantly alter the efficiency ranking of models with chillers. Thus, for the purposes of calculating the energy standards lines within this report the continued use of W = 0.75 is recommended. The issue in points b) and d) of whether models can be properly discriminated within the GEA database is disingenuous. At the time of their study the GEA circulated an extensive questionnaire to manufacturers requesting their assistance in supplying data on the energy consumption of their models and canvassing their opinions on product categorisation. It appears that in a small number of cases insufficient information may have been supplied for models to always be placed in the correct category. The eco-labelling research team are attempting to correct the GEA database and have been supplied with an extra quantity of Italian manufacturers data. It is relevant that the eco-labelling study is sponsored by the Italian manufacturers association ANIE who produce almost half of all refrigeration models sold in Europe. The various manufacturers associations were fully aware of the GEA study and had the opportunity to draw attention to deficiencies in the database long before the report was released. The fact that they did not raise these points earlier suggests that they may have wished to undermine its findings. Despite these observations it appears unlikely that the revisions to the database will make a substantial difference to the assessment of the energy reference lines. This is illustrated in Section 5.1.2 where the eco-labelling study energy reference lines are compared with the GEA reference lines. The draft eco-label study proposed the following product categorisation: 01 Fridges without a low temperature compartment 02 Fridges with 0 star compartment 03 Fridges with 1 and 2 stars compartments 04 Fridges with 3 stars compartment 05 Fridge-freezer with double doors, 4 stars 06 Fridge-freezer with double doors, 4 stars, no-frost 07 Fridge-freezers with more than two doors, 4 stars 08 Fridge-freezers with more than two doors, 4 stars, no-frost 09 Upright freezers 10 Upright freezers, no-frost 11 Chest freezers 12 Chest freezers, no-frost These categories are the same as the GEA's except for: ù making explicit categories for no-frost models, ù distinguishing between fridge-freezers on the basis of the number of doors, ù combining 1 and 2 star fridges into the same category, ù distinguishing between 0 star fridges and fridges without a low temperature compartment. The issue of separate categories for no-frost models is discussed in Section 5.3.2 on short-term efficiency standards. Splitting fridge-freezers on the basis of whether they have two or more doors is only a concern for a small number of models which have a "preserver" compartment (one that operates at about 0oC to 3oC). Two door models usually have only one compressor, while 3 or more door models often need two compressors to be able to control the extra compartments temperature. The GEA investigated this issue (Michel 1992) and concluded that although models with varying numbers of compressors have significantly different adjusted volumes and energy consumptions, the energy reference lines generated from each group are not significantly different. For this reason they concluded there was no justification for splitting the fridge-freezers in to more than one group. The GEA conducted full statistical tests to demonstrate that there was a significant difference in the energy reference lines of 1 star and 2 star fridges. They also found that fridges with chillers and fridges with no frozen food compartment did not need to be maintained as separate groups because they gave statistically indistinguishable reference lines. The eco-labelling study presented no extra analysis to support their product groupings and therefore the GEA groupings, which are also accepted for the energy-labelling Directive, would appear to be justified. For the standards discussed and proposed in this report we accept the GEA product groupings as being appropriate. 4: REFRIGERATOR DESIGN ISSUES 4.1 DESIGN CONCEPTS A refrigerator system can comprise one or more cooling systems and one or more cooled compartments. The energy required by the system is determined by the size of the thermal load in each cooled compartment (i.e. how many Watts of heat are to be extracted) and the efficiency with which the cooling system(s) satisfy the thermal load(s). 4.1.1 COOLING SYSTEM OPERATING PRINCIPLE The cooling system is a mechanical heat engine which pumps a working fluid (the refrigerant) around a circuit, see Figure 4.1. As the refrigerant is circulated it is cyclically compressed and expanded, releasing and absorbing heat in the process. More specifically, the refrigerant enters the compressor as a low temperature, low pressure vapour where it undergoes adiabatic compression and exits as a high temperature, high pressure vapour. It is then fed to the condenser, which is in good thermal contact with the ambient, causing the vapour to lose heat and condense into a lower temperature, high pressure liquid. From here the liquid is fed at high pressure into a storage chamber before undergoing an adiabatic expansion by passing through a throttling valve. When a liquid passes adiabatically through a narrow opening (a needle valve) from a region of constant high pressure to a region of constant lower pressure, it is said to undergo a "throttling process". If the liquid is "saturated" the throttling process produces cooling and partial vaporisation. This low temperature, low pressure, liquid-vapour mixture is fed into the evaporator which is a heat-exchanger in good thermal contact with the inside of the refrigerator cabinet. The mixture absorbs heat from the cabinet which causes it to become completely vaporised but only slightly raises it's temperature. The low-pressure, low temperature vapour is fed once again into the compressor and the cycle is repeated. The evaporator and condenser units are both heat exchangers. The evaporation stage is arranged to occur within the refrigerator cabinet and the condensation stage to occur away from the cabinet, so that heat is "pumped" from the cabinet to the outside and the inner cabinet temperature is lowered. The other main component of the cooling system is the compressor. Most existing refrigerators use reciprocating compressors which employ a standard vertical piston arrangement to compress the refrigerant. Rotary compressors use an alternative approach, employing a rolling piston of similar design to rotary Wankel engines to compress the refrigerant vapour. Rotary compressors have the advantage that the fluid flow is continuous and pressure is more constant which lowers unwanted heat generation through hysteresis. 4.1.2 CABINET DESIGN AND THERMAL LOADS The cabinet contains the evaporator and supports the condensing unit, as well as providing the storage space for foodstuffs. The cabinet's outer shell is usually made of pressed steel with spot welded seams. The inner shell is made of an impermeable plastic. Foam thermal insulation is blown in-between the two shells until the space is completely filled. Once set, the foam helps to add structural rigidity as well as providing insulation. The size of the thermal load for a given operational temperature is principally determined by the quality of the insulation used to thermally separate the cooled space from the ambient. Typical modern models use 30mm of polyurethane foam (PUR) insulation in the outward facing sides of the fridge compartment and 50mm for the outward facing sides of the frozen food compartment. Fridge doors typically use 20mm of PUR insulation, while freezer doors might use 50mm. A typical chest freezer cabinet uses 60mm of PUR insulation. The thermal conductivity of polyurethane foam (PUR) blown with HFC-134a is about 0.024 W/mK. This is about 35% lower than the mineral wool insulation that was favoured until the sixties by manufacturers in Europe, and until the 1973 oil shock by manufacturers in the United States. 4.1.3 DEFROSTING SYSTEMS When the evaporator temperature is sub-zero water vapour condenses out of the air on to the cold surface of the evaporator plate and freezes. The resulting ice build up reduces the storage space in the refrigerator, increases the thermal resistance of heat transfer between the evaporator and the cabinet, and reduces air circulation from the evaporator to the rest of the cabinet. This makes it necessary to defrost the evaporator on a regular basis. There are four categories of defrosting system: ù manual, ù semi-automatic, ù auto-defrost, ù no-frost systems. The first three categories simply refer to the amount of user involvement, such that with manual systems the user has to start and end the defrosting process, with semi-automatic systems the user only has to initiate the defrosting process, and with auto-defrost systems the whole process is automatic. Defrosting methods may be passive or active. With passive defrosting the cooling cycle is stopped until the ice thaws from the evaporator. Active systems either utilise an electric heater, or a hot gas system which uses the heat in the vapour compressor discharge line and the condenser to defrost the evaporator. In no-frost (sometimes called "frost-free") systems the evaporator is located outside the refrigerator compartment. On the running part of the cycle, air is drawn over the evaporator and is blown into the freezer and fridge compartments using an electric fan. During the off part of the cycle the evaporator defrosts automatically. The melt water is taken to a pan above the compressor where the heat causes it to evaporate and return to the atmosphere. 4.1.4 IMPACT OF OZONE DEPLETION ON CHOICES OF REFRIGERANT AND FOAM BLOWING AGENTS Until the discovery of the hole in the ozone layer above the Antarctic in 1985 and the subsequent signing of the Montreal Protocol to control chloroflurocarbons production (CFCs) in 1987, manufacturers were using CFCs for both the refrigerant and as the insulation foam blowing agent. This class of chemicals had the advantages of good thermodynamic properties, low toxicity and being relatively inert, but were found to be contributing to the thinning of the ozone layer. CFC-12 was the favoured refrigerant for domestic refrigerators while CFC-11 was used as the blowing agent. The Montreal Protocol and the subsequent amendments have forced manufacturers to find other refrigerants with less or zero ozone depleting potential (ODP). Originally manufacturers claimed that switching to zero ODP refrigerants was not possible without incurring a significant energy efficiency penalty. This is now known to be un-true. Within Europe two alternative classes of refrigerant are currently being used, HFCs and hydrocarbons. HFCs, such as HFC-134a, have been adopted by many manufacturers in their latest model ranges, but although these gases have zero ODP they are very potent greenhouse gases. The contribution of HFCs to the Total Equivalent Warming Impact (TEWI) over the lifespan of a typical European refrigerator is investigated in Section 10. The most environmentally benign alternatives are the hydrocarbons, such as propane-butane mixtures or iso-butane. Recent changes in EU health and safety law mean that manufacturers can now use up to a kilogram of propane, butane, or their mixture within a domestic refrigerator cooling system. Propane-butane refrigerants were pioneered in modern times by the east German company DKK Scharfenstein after Greenpeace approached the German government and encouraged them to put up the development money. This led to the production of the Greenfreeze refrigerator which became available on the European market in early 1993; the first domestic vapour-compression refrigerator in modern times to be CFC, HCFC and HFC free. Since this time Bosch-Siemens, Liebherr and Miele have also released models using propane-butane as the refrigerant. The Greenfreeze uses only 24 grammes of propane-butane as the refrigerant, which tests have shown poses no greater safety risks in use than HFC or CFC refrigerants. Propane-butane has excellent thermodynamic and heat transfer properties, is cheaper and has been shown to perform as well in actual refrigeration systems as CFCs and HFCs. According to John Missenden of South Bank University "there is a consensus of informed opinion that if energy efficiency is an important goal, then hydrocarbons such as propane are the right way to go for small scale refrigeration" (Missendon & James 1992). A report by Star Refrigeration of Glasgow stated "Propane will eventually be adopted as the preferred refrigerant for small sealed systems because it is the most efficient and most benign substance available and because it can be used with negligible risk" (Forbes Pearson 1991). Similar developments have caused manufacturers to seek substitute foam blowing agents to CFC-11. Of the 1993 model releases manufacturers appear to be opting either for HFC-134a or pentane. Pentane has advantages because it is non-toxic, has insulation equivalence to CFC-11, has zero ODP, zero global warming potential (GWP) and is cheap to supply. Its only disadvantage is it's flammability, which does not present a significant safety problem in use, but means that a manufacturer has to invest in additional production safety features. Liebherr made a decision to switch to pentane for their blowing agent in September 1992, with the aim of blowing all their foam using pentane by March 1993. Other companies, especially in Germany, are following suit. In Liebherr's assessment of alternative foam blowing agents they concluded that by 1994 all European manufacturers would have switched to pentane. 4.1.5 MIXED FRIDGES AND FREEZERS A large proportion of refrigerator models have a mixture of a fridge and either a freezer or a frozen food compartment. These hybrid systems complicate the arrangement of the cooling system because two different compartmental operating temperatures are required. For a small number of fridge-freezers a separate cooling system is used for each compartment, which has the advantage of allowing thermodynamic optimisation of the individual cooling circuits. However, the component duplication required in this solution is expensive and is often only justified for large, top of the range, combination models. For the remaining models there are a variety of solutions involving the use of one primary cooling system: With `air spillover' systems the evaporator is located at the top of the cabinet in the freezer or frozen food compartment. The temperature in the fridge compartment is controlled by regulating convective air circulation from the evaporator to the fridge below. `Secondary refrigerant systems' have the primary evaporator in the freezer or frozen food compartment and use a secondary heat exchanger to absorb heat from the fridge compartment. The refrigerant from the secondary system is transferred into the evaporator of the primary system which has the capacity to manage both heat loads. There are two types of `combination evaporator systems', i.e. systems with two evaporators but one compressor and condenser. One maintains the same pressure and temperature in both evaporators but controls the temperatures in each compartment by the size and thermal capacity of each evaporator. The other controls temperature by linking the two evaporators with a pressure reducer. This maintains a higher pressure and hence temperature in the fridge's evaporator than the freezer's. 4.2 THE TECHNICAL POTENTIAL FOR IMPROVING REFRIGERATOR ENERGY EFFICIENCY It is shown in Section 5.2 that the current stock of refrigerators have not reached any technical energy efficiency limit. It is important to define present technical limits and how they may change in the future for sensible regulatory decisions to be made. The technical limits should not be confined to physical phenomena but should also encompass implementation costs. Fundamentally there are two areas in which the energy efficiency of a refrigerator can be improved: ù reducing the thermal load experienced by the cooling system, ù improving the efficiency of the cooling system. 4.2.1 IMPROVING THE EFFICIENCY OF THE COOLING SYSTEM Domestic refrigerators involve two energy conversion stages. The first is the conversion of electrical energy into mechanical energy within the compressor, and the second is the conversion of mechanical energy into thermal energy through the refrigeration cycle. Whenever an energy conversion occurs there is an energy loss and an associated system efficiency. The efficiency of the whole cooling system is defined in terms of the amount of heat extracted for a given expenditure of electricity. If these quantities are expressed as energy per second (Watts) then the cooling system efficiency, known as the Coefficient Of Performance, is: C.O.P = refrigeration capacity/net power input [4.1] The higher the C.O.P. the more efficient the cooling system. For a perfect (lossless) cooling system the theoretical maximum possible C.O.P. can be defined thermodynamically in terms of the system temperatures as: C.O.P.max = T2/T1 - T2 [4.2] where T1 is the temperature (K) of the evaporator and T2 the temperature of the condenser (K). For typical domestic refrigerators, the C.O.P. is between 1.0 and 2.0 (with higher values occurring for systems with higher refrigeration capacities), which is still a long way short of the theoretical limits. It is clear from equation [4.2] that if the evaporator and condenser temperatures are similar the efficiency of the cooling system improves. For this reason, good ventilation of the condenser is important. It has been proposed that north European kitchens could be modified to allow the back of the refrigerator (the side with the condenser) to be exposed to air outside the building and hence to cooler conditions than inside the kitchen (Norg†rd 1989). For the same reason it is also important that the evaporator is in good thermal contact with the whole of the cooled space, otherwise a locally over cooled micro-climate occurs and the C.O.P. drops. Consequently, the build up of frost on the evaporator plate is undesirable because it insulates the heat exchanger from the rest of the cabinet. The need to circulate heat between the whole of the cooled space and the evaporator has a number of design implications. It dictates that frozen food compartments should be sited at the top of units with a simultaneous fridge compartment so that cold air from the evaporator will sink and warmer air from the fridge rise. These convecting air currents help keep the temperature distribution within the cooled spaces more even and reduce frost build-up round the evaporator. Were the frozen food compartment sited beneath the fridge the cold air would stagnate at the bottom, causing temperature stratification within the cabinet and increasing the ice build-up on the evaporator. There are three design options to reduce the temperature difference between the evaporator and condenser heat exchangers: ù improve ventilation of the heat exchangers, ù enlarge and/or redesign their surfaces to facilitate the rate of heat transfer, ù increase their thermal capacity to extend the time available for heat transfer. Combinations of fridge and freezer compartments within one refrigerator unit pose problems for the cooling system, because it is thermodynamically inefficient to use the same cycle to attain the 5oC fridge temperature and the -18oC freezer temperature. This is particularly the case if one evaporator serves both fridge and freezer, especially because it experiences moisture in the air from both. As the fridge will be used much more frequently than the freezer this greatly increases the exposure of the freezer evaporator to air at room temperature and humidity and therefore greatly increases the rate of frost build-up. The performance of the compressor is the other main determinant of the overall cooling system efficiency. In recent times a number of promising technologies have become available some of which are close to commercialisation. Sun Power Incorporated (1992) have developed a prototype compressor which claims an energy efficiency ratio (EER) of 6.2 with the potential to reach 6.8 with further design improvements compared to EER's from under 4.0 to 5.5 for existing models. An EER of 6.2 corresponds to a C.O.P. of 2.0 with a heat removal of 250W from -23oC to 32oC. Conventional, reciprocating compressors use a crank mechanism to convert the rotational motion of an electric motor into a sinusoidal up and down motion, which drives a piston and pumps the refrigerant vapour. This type of motion requires lubrication and loses a significant amount of energy through friction. The Sun Power technology uses a linear motor that drives the piston up and down without needing a crank shaft. The piston resonates on a spring which helps reduce the motor workload. Frictional forces are much reduced through the use of linear power and gas bearings with a long clearance seal between the cylinder and the wall. The designers claim these features result in a 9% to 15% energy efficiency improvement over current state of the art compressors. The linear compressor system has a constant EER as a function of cooling capacity while the EER of reciprocating compressors increases with cooling capacity. Thus, the Sun Power design would give greater efficiency gains for well insulated refrigerators with low thermal loads. The use of gas bearings eliminates the need for oil lubrication of the piston/cylinder arrangement which avoids problems concerning the compatibility of some refrigerants and the traditional lubrication oils. Consequently, the designers claim that the compressor is compatible with any refrigerant that doesn't chemically attack the apparatus. They also claim that between 10% and 25% of additional energy efficiency improvements can be gained, depending on application, through the ability to continuously vary the pumping capacity to match the instantaneous cooling load. Most conventional compressors cycle on and off from a single pumping capacity, which increases motor losses and noise, and reduces temperature control. Sun Power Inc. also claim that their compressor would be cheaper to manufacture than traditional models because it uses less materials. Thus, it appears that this technology could save between 18% and 36% of refrigerator energy consumption and for less cost than the existing models. Commercialisation of the product is under way with Americold, a US compressor manufacturer associated with Electrolux, and Bosch-Siemens likely to be involved. It is thought models will be on the market in 1995 to 1996. In addition to the Sun Power Inc. system there are a number of other promising compressor technologies. For instance, Americold has evaluated Lorentz, La Brecque, Magnetic Resonance, and Magneto-caloric cooling systems and "believes any of these have a better chance at ultimate success than the pulse tube in replacing gas cycle in household refrigerators" (Simpson 1993). The pulse tube is another alternative compressor technology which uses electrical power to produce resonant sound waves with pressure amplitudes up to 100lb per sq in. These sound waves compress the gas and drive the refrigeration cycle. For existing refrigerator models it is possible to reduce the compressor motor losses associated with on/off cycling simply by replacing the conventional plug with a special controller plug (McNaught 1993). At full load the standard compressor induction motor used in refrigerators is at it's most efficient, but it's efficiency drops dramatically as the motor load falls. This is because two of the three sources of energy loss in the motor, the "iron" losses and the friction losses, are constant under all load conditions. Iron losses are greater than friction losses and result from unwanted magnetic eddy currents generated in the metal structure of the motor. The motor is only at full load at the beginning of each on cycle while the refrigerant reaches its normal running pressure and temperature. After two to four minutes the motor runs at only 50-60% of its full load. The motor controller plug works by reducing the voltage supplied to the motor at part load, this reduces the iron losses because they are proportional to the square of the voltage. The motor load is monitored by the controller from its back emf. At half load the controller cuts the average voltage across the motor by 20%. Trials on over thirty refrigerator appliances demonstrated average refrigerator energy savings of 7.4% with maximum savings of 27%. However, further improvements in the controller have since been made which are expected to improve its performance. The UK energy Efficiency Office have part funded a development project with Savawatt (UK) Ltd to produce a domestic version of the controller plug. 4.2.2 IMPROVING THE EFFICIENCY OF REFRIGERATOR INSULATION The most cost effective way to reduce refrigerator energy consumption is to reduce the thermal load (and hence electrical load) through better insulation of the cooled space. According to Benson and Potter (1992): "Contrary to popular opinion, the energy use of a household refrigerator is not closely related to the frequency of door openings or the amount of food kept inside. Rather even with improvements in insulation over the past 20 years, between 70% and 95% of the electrical load, may still be attributed to the thermal performance of the insulated shell. The remaining load is caused by gaskets, which also have great improvement potential, and from (automatic) defrost heaters and antisweat heaters (which are not so common in European models)." Heat loss through the door seal gasket can be a significant component of the overall thermal load. Close fitting doors with a rebate can help reduce this load as can the calibre of gasket insulation. For the GEA analysis rubber/magnet gaskets were assumed with conductivity of 0.04W/mK. However, seals with conductivity of 0.03W/mK or better are available. The most obvious way to improve the insulative performance of the cabinet and the doors is to increase the insulation thickness. However, this will either reduce the internal volume of the refrigerator or increase it's external dimensions. The GEA tackled this problem, as discussed in detail in Section 6.3, by ascribing a penalty cost for lost storage space as the insulation is increased, in this manner they were able to estimate an optimum insulation thickness which corresponds to the least life cycle cost for the consumer. 4.2.3 VACUUM INSULATING PANELS (VIPS) The alternative means of reducing the thermal load is to use a better quality insulant. Traditionally the cabinet and door insulation has been provided by blown foams such as polyurethane. However, recently evacuated panel technologies (vacuum panels) have been developed which offer the opportunity for greatly improved energy performances and a better all round product for the consumer. Vacuum panels operate on the same principle as a thermos flask, in that through creating a vacuum heat can only be transferred by radiation between the inner and outer surfaces and by conduction at joining edges. Convective heat transfer is eliminated and therefore the overall quality of the insulant is improved. There are 3 principal types of vacuum insulating panels under development: ù encapsulated foam panels ù precipative silica filled panels ù fibreglass panels. Encapsulated foam panels are being developed by ICI in Belgium in conjunction with 6 major refrigerator manufacturers and consist of a foam structure with open cells sealed in a vacuum. The conductivity of the panels is expected to be about 0.005 to 0.007W/mK. The longevity of the panels depends upon the encapsulation method, and is still being assessed. It is thought that full scale commercial production could begin in less than three years (Jeffs 1993). This technology is the least mature of the three, but may result in the cheapest, lightest panels. Silica powder filled panels have been developed by Degussa of Germany and by General Electric in the U.S. The Degussa product comprises a filler of highly structured silica (SiFiO2) which is compressed to form a board before being evacuated and sealed between gas and water tight barrier films. These panels have a thermal conductivity of 0.006W/mK to 0.008W/mK, equivalent to R-20 to R-22 insulation and operate with an internal pressure of 1 to 2 mbar (Strack 1993). Conventional PUR foam has a thermal conductivity of 0.024 W/mK, thus, the Degussa panels are between 3 and 4 times more efficient insulants. The Degussa panels have already been included in a commercially available refrigerator, the AEG 275 GSJ_VIP freezer, see Figure 4.2. During manufacturing the panels are attached to the outer metal casing, before the remaining space between the inner and outer casing is filled with PUR foam. The PUR foam is used primarily to provide structural rigidity. Unfortunately, because there were only two sizes of vacuum panel available, the panels do not cover the whole surface area of the cabinet leaving substantial gaps where there is only PUR foam insulation. This allows thermal bridging to occur and lowers the overall insulation quality. Nonetheless, tests have shown that the model with the panels, AEG 275 GSJ-VIP, used only 81% of the energy of the identical model without panels, AEG 275 GSJ. On other test models of fridges, fridge-freezers and freezers, Degussa have found energy savings of between 20% and 25% (Reuter & Sextl 1993a). Degussa estimate the cost of their panels to be between 50 to 80 DM per sq metre per 25mm thickness, depending on the size of the panel. This equates to a cost in the range of $2.50 to $3.80 per square foot for 1 inch thick panels. The longevity of vacuum panels has been a primary issue influencing their usage. Degussa are currently testing the durability of their panels with the German refrigerator manufacturers association ZVIE. The results show that the panels have a maximum reduction in effectiveness of 25% after 15 years i.e. increasing their conductivity from 0.006W/mK to 0.008W/mK (Reuter & Sextl 1993b). As Degussa are the World's largest supplier of silica, they are offering free production franchises to all who wish to manufacture these panels. Electrolux Siegen, the German subsidiary of the Swedish Electrolux AB group, will operate the first large scale VIP production plant starting from mid 1994 with a capacity of more than a million panels per year. The third class of VIPs are the fibreglass panels pioneered by the American engineering company Owens-Corning of Toldeo, Ohio. The construction consists of a hermetically sealed steel outer casing with an evacuated thermally designed fibreglass infill. Owens-Corning were claiming to have a minimum insulation of R-50, equivalent to a conductivity of 0.0029W/mK (Remich 1993, Arch 1993). Recently they have tested 6-ft2 panels with insulation in excess of R-100 at the centre but with effective panel insulation of R-60 because of edge losses (Arch 1994). The panels retain their performance at very high temperatures which means they have insulant applications for ovens and water heaters as well as refrigerators. Owens-Corning have teamed up with the refrigerator manufacturer Maytag and are investigating using the panels in refrigerator cabinets. The structure of these panels is sufficiently strong for the cabinet to be constructed from the panels alone, thereby avoiding the need for the panels to be inserted inside a conventional foam filled cabinet. The integrated cabinets have the advantage that the panel coverage of the cooled space is total apart from necessary intrusions such as thermostats, lights, the evaporator and gaskets. This will greatly improve the actual energy savings achieved. Projected costs of the Owens-Corning panels are thought to be between $5 and $7 per square foot per inch thickness, although these are expected to fall with time (Arch 1993). The longevity of the panels is projected to be between 20 to 22 years without significant deterioration. Owens-Corning claim this is a much better performance than precipative silica panels (Arch 1993). The US DOE have conducted a vacuum panel performance study under which vacuum panel producers were invited to submit samples of their products to the agency for independent testing and evaluation. The results have been encouraging with effective panel insulations of R-35 being attained at projected costs of $1.40 per square foot per inch. These costs are forecast to drop below $1.00 by 2010 (Colley 1993). One of the most detailed studies on the cost of manufacturing precipitative silica VIPs and of incorporating them in the refrigerator cabinet was produced by the US EPA (Fines 1992). Every aspect of the production and assembly process was considered from the materials, plant design, and annual operating costs of the panel plant, to the refrigerator assembly costs. The costs assumed production of 300,000 refrigerators a year. The aggregated values are shown in Table 4.1 and give a final panel cost of $1.39 per square foot per inch for a 21 cubic foot fridge. On top of this may need to be added non recurring program costs which cover the planning and implementation of the project from start-up. These costs only have to be paid once and amount to $0.54 per square foot and inch thick panel if a five year operation period is assumed. In fact, these costs are likely to be spread over more than five years, so a total panel cost of $1.93 can be viewed as an upper limit. The high insulative performance of VIPs gives them additional advantages over blown foam: they are more compact, they maintain a much more stable temperature throughout the cabinet which enables much better food preservation properties, they reduce the build up of condensation (thereby saving on defrost time and the energy required to do it). Furthermore, in the future it is possible that consumers could buy the electricity for refrigerators at a reduced rate because vacuum panels would allow the use of off-peak electricity and the flattening of the domestic demand profile (this offers a double saving because the temperature of the kitchen will be cooler during the off-peak charging periods). 4.2.4 SIMULATED REFRIGERATOR ENERGY CONSUMPTION USING VIPS To enable the effect of different design options to be explored without the need to construct physical prototypes it has become common practice to use computer simulations for what is known as an "engineering analysis". One such refrigerator simulation program is the SPADE model, developed by the Refrigeration Laboratory of the Technical University of Denmark (dk-TEKNIK) and used by the GEA in their study. In a study by dk-TEKNIK (Pederson 1992) the electricity consumption of a number of SPADE simulations of existing fridges, freezers, and fridge-freezers were compared to the measured values. The results showed that the average difference was only 2.7% with a maximum value of 6.5%, which compares favourably with other simulation models. To explore the energy benefits of using vacuum panel insulation Greenpeace contracted the Dutch Institute of Environmental and Energy Technology, TNO, to conduct an engineering analysis with SPADE. For the simulations to be most meaningful it is preferable if the design alterations are made to a standard or "average" design of refrigerator known as a base-case model. The GEA performed their own refrigerator simulations using the SPADE program, described in Section 6, for which they first established a base-case model for each category of refrigerator listed in Table 3.1. The base-case models were arrived at after examining the models on the market, holding discussions with manufacturers and circulating an extensive questionnaire requesting design details to manufacturers. These base-case designs were also used by TNO for the Greenpeace analysis. Two categories of refrigerator were examined: fridges without a frozen food compartment (R6) and upright freezers (F2). For the R6 engineering analysis the Owens-Corning VIP concept was assessed. It was assumed that cabinet design was integrated, i.e. that the whole cabinet including the door was made of the VIP construction excepting the fittings. The insulative quality of the VIPs was taken to be 0.0029W/mK (R-50) which is conservative. Fixture intrusions into the VIP constructed cabinet were modelled, such as the evaporator, light, thermostat etc., while standard door seals were assumed. In all other respects than the vacuum insulation the model examined was identical to an average European fridge with no frozen food compartment defined by the GEA R6 base-case. The energy consumption results of successive simulations using progressively thicker Owens-Corning type VIPs are shown in Figure 4.3. The GEA base-case model, using foam insulation, is marked with a cross. These simulations show the extraordinary potential of VIP technology to improve refrigerator energy efficiency. For the same thickness of insulation as used in the base-case model the VIP model uses a sixth (-83%) of the energy! This result is even more remarkable given that no other design improvements were assumed. The asymptotic shape of the energy consumption curve reflects this, because as the cabinet insulation is systematically improved the heat loss through the door seals and fixture intrusions becomes relatively more important. For the F2 category engineering analysis the VIP characteristics cited by the USDOE study were assumed, i.e. a conductivity of 0.004W/mK (R-35). Once again it was assumed that the panels covered the whole cabinet surface except for the fixture intrusions. However, with this design the cabinet was not constructed from the VIPs, rather it was assumed that the VIPs were attached to the inside of the cabinet outer steel casing and that 0.5 inches of foam insulation were injected into the space between the inner cabinet lining and the VIPs to provide structural support. In all other respects than the vacuum insulation the model examined was identical to an average European upright freezer defined by the GEA F2 base-case. The energy consumption from using progressively thicker VIPs but with a fixed thickness of foam is shown in Figure 4.4. Once again the results are impressive, with the VIPs giving a 68% reduction in energy consumption for the same thickness of total insulation as the base-case model. This is less than the R6 category simulation because the conductivity assumed for the vacuum panels is higher. 5: SHORT-TERM ENERGY STANDARDS 5.1 ENERGY REFERENCE LINES 5.1.1 GEA REFERENCE LINES The GEA produced energy reference lines for each of the refrigerator groupings they identified. These lines give the statistically averaged dependency of a given category of refrigerator appliance on its adjusted volume and provide a base line for establishing appliance efficiency. The GEA found that there was no statistically significant difference in the energy reference lines for categories R5 and R6, and that therefore it was sensible to combine them in to one category with a single reference line. The results are shown graphically in Figures 5.1 - 5.7, while the equations defining the energy reference lines are given in Table 5.4. 5.1.2 COMPARISON OF GEA AND ECO-LABELLING STUDY ENERGY REFERENCE LINES Since the release of the GEA study, Benoit Lebot of the GEA has analysed the ENEA eco-labelling study energy-reference lines and compared them to the GEA lines. The results are shown graphically by category in Figures 5.8 - 5.15 where: ù the line "ENEA Italy" is the eco-label study energy reference line produced using a database of 644 models manufactured in Italy and supplied by the Italian manufacturers association, ANIE, ù the line "ENEA Europe" is the energy reference line produced using the GEA database of refrigerator models from all Europe combined with the extra Italian data, ù the line "GEA" is the GEA energy reference line. All three reference lines are seen to be very close for categories: R5 (fridge with a 0-star frozen food compartment) and ENEA 02, R6 (fridge without a frozen food compartment) and ENEA 01. This result appears to confirm the GEA assertion that it is reasonable to represent these two categories with one energy reference line. For fridges with 1 and 2 stars (R1, R2 and ENEA 03) the reference lines show a slight divergence, but this is explained by the grouping of the 1 and 2 star fridges into a single category for the eco-labelling study. For chest freezers (F1 and ENEA 011), upright freezers (F2 and ENEA 09), and fridges with a 3-star frozen food compartment (R3 and ENEA 04) the reference lines for the all-Europe databases are very close. However, the Italian data gives quite different results for upright freezers and fridges with a 3-star frozen food compartment, but fairly close agreement for chest freezers. The greatest divergence occurs for the fridge-freezer categories (R4 and ENEA 05). The discrepancies in the energy reference lines may be the result of some significant differences between the Italian produced models and those produced elsewhere in Europe, or they may be a statistical artifact caused by the limited quantity of the Italian data by category. However, where divergences between the GEA database reference line and the Italian database reference line are most evident (categories R3 & ENEA 04, R4 & ENEA 05 and F2 & ENEA 09) a consistent pattern has emerged. The Italian data fits to lines with shallower gradients and larger y-axis intercept values than the GEA data. Within the eco-labelling study the energy consumption of no-frost models of fridge-freezer (category ENEA 06) were more dependent on adjusted volume than the standard models (category ENEA 05). Consequently, the shallower gradient of the Italian data is consistent with the eco-labelling study assertion that the GEA database is a mixture of no-frost models and standard models. Whatever the cause, the discrepancies between the reference lines are not too serious over the range of adjusted volumes actually experienced. Benoit Lebot compared the values given by the GEA and ENEA Europe energy reference lines for the adjusted volumes of a small data-set of 36 refrigerator models and found that the maximum difference was 6.9%. 5.2 DISTRIBUTION OF REFRIGERATOR ENERGY EFFICIENCY 5.2.1 DISTRIBUTION OF ENERGY EFFICIENCY BY NATIONAL MARKET The frequency distribution of all refrigeration appliance energy-efficiency indices by country have been computed from the GEA database and are shown in Figures 5.16 - 5.23. In order to illustrate the random nature of the distribution of efficiency indices a Gaussian distribution has been fitted to each national distribution. In general it appears that the actual distributions are a little more peaked than a Gaussian and with a slight tendency toward an earlier cutoff point for the more efficient left-hand side than the less efficient right-hand side. The significant aspect of these graphs is that they show clearly that current refrigeration models have not encountered any physical limit to energy efficiency. This follows because were there such a limit the distributions would be negatively skewed (i.e. appreciably more models with an efficiency index less than the mean value than those greater). The evenness of the current distributions is especially remarkable given that the energy efficiency index is not an open-ended parameter (it has a lower limit of zero but an upper limit of infinity) therefore one would have anticipated models to have a strongly negatively skewed distribution. The fact that the distributions are so even proves that substantial energy efficiency improvement is readily attainable. The contrast in the efficiency of models sold between national markets is marked. The average efficiency index of models sold in Portugal is 1.28 while in Denmark it is 0.94. Of the larger countries, the German market is appreciably more efficient than the others, while the UK and Italian markets are least efficient. The energy efficiency distribution by country is summarised in Table 5.1. TABLE 5.1: REFRIGERATOR ENERGY-EFFICIENCY DISTRIBUTION STATISTICS BY NATIONAL MARKET (FOR ALL CATEGORIES)*1 Energy Efficiency Index Country Number of Models Mean Maximum Minimum Standard Deviation Denmark 402 .94 1.54 .35 .19 Germany 959 .95 2.56 .43 .21 Spain 173 1.01 2.25 .43 .22 France 655 1.04 1.83 .43 .19 Italy 425 1.07 2.22 .65 .23 Netherlands 996 .99 1.74 .38 .18 Portugal 62 1.28 3.41 .20 .46 UK 72 1.08 1.83 .54 .24 *1 The data is from the GEA database. 5.2.2 DISTRIBUTION OF ENERGY EFFICIENCY BY PRODUCT CATEGORY Figures 5.24 to 5.30 show the frequency distribution of refrigerator energy efficiency indices by refrigerator category; also shown is the best Gaussian fit. Again the distributions are surprisingly even and mostly uni-modal which indicates that the product categorisation is sound. In particular, there is no evidence of bi-modal behaviour for the fridge-freezer (R4) and upright freezer (F2) categories which suggests that no-frost models were not included in these data-sets. The exception is the chest freezers (F1) that have a bi- or even tri-modal distribution which seems to indicate there are at least two broad types of design within the category. However, this does not necessarily mean that the categorisation is inappropriate because there is no evidence that the seemingly distinct design-types effect any other aspect of the chest freezers utility than the energy consumption. The most efficient models in the GEA database are listed by category in Table 5.2. These models are up to 65% more efficient than the average, illustrating that were similar design practices applied for all models the energy consumption of new refrigerators would be about 40% of the present average. It is notable that Danish and German manufacturers are producing the most energy-efficient models in all categories. TABLE 5.2: THE MOST ENERGY EFFICIENT MODELS IN THE GEA DATABASE BY CATEGORY Category Model Make Energy Efficiency Adjusted Energy & Mark Index (EEFF) Volume Consumption (litres) (kWh/yr) R1 Vibo Cold R293 0.75 84 183 R2 Frigor K 42 mini 0.41 45 109 R3 Liebherr KT 1483 0.55 148 183 R4 Vestfrost SKF 375 0.52 411 321 R5 & R6 Gram LER 200 0.35 196 102 F1 Liebherr GTS3663 0.43 744 219 F2 Liebherr GSS2663 0.44 452 219 5.2.3 DISTRIBUTION OF ENERGY EFFICIENCY BY MANUFACTURER By applying the energy efficiency index of equation [3.3] to the models in the GEA database and aggregating the results for each manufacturer it is possible to rank the manufacturers according to the energy efficiency of their entire range of brands. The distributions of model energy efficiency for the top twelve manufacturers are shown graphically in Figures 5.31 to 5.42, while the average energy efficiency ranking of all the major manufacturers is given in Table 5.3. The top eight ranking companies are German or Danish manufacturers, while the bottom seven are either UK and Italian companies or companies specialising in East European models. Of the top three manufacturing conglomerates Bosch-Siemens perform best, while Electrolux and Whirlpool both produce models that are less efficient on average than the norm. TABLE 5.3: ENERGY EFFICIENCY RANKING OF REFRIGERATOR MANUFACTURERS *1 Manufacturer Rank Number Energy Efficiency Index of Models Mean Maximum Minimum Standard Deviation Gaggenau*2 1 25 .85 1.12 0.70 0.12 Liebherr 2 213 .89 1.61 0.43 0.20 Oetker- Gruppa *3 3 12 .90 1.13 0.73 0.14 AEG 4 158 0.91 1.43 0.43 0.19 Derby 5 8 0.92 1.31 0.43 0.19 Miele 6 116 0.93 1.43 0.66 0.13 Gram 6 77 0.94 1.37 0.35 0.18 Bosch Siemens 8 554 0.98 1.52 0.43 0.17 Candy 9 119 0.98 1.72 0.68 0.28 Blomberg 10 84 0.98 1.31 0.71 0.12 Fagor 11 38 1.01 1.74 0.58 0.21 Quelle *4 12 88 1.02 1.70 0.50 0.25 Thomson 13 210 1.03 1.62 0.63 0.15 Maytag 14 5 1.03 1.09 0.99 0.05 Kuppersbusch*4 14 65 1.03 1.31 0.80 0.12 Electrolux 14 755 1.03 1.89 0.46 0.21 Whirlpool 17 509 1.04 2.22 0.51 0.20 Merloni 18 129 1.05 1.74 0.65 0.21 Gorenje 19 101 1.06 2.56 0.72 0.22 Sidex 20 15 1.09 1.38 0.81 0.15 Ocean 20 68 1.09 1.83 0.73 0.21 GEC 22 25 1.15 1.35 0.83 0.13 DKK Sharfenstein*5 23 5 1.20 1.32 1.08 0.10 Lec 24 14 1.46 1.83 1.08 0.24 *1 The data is from the GEA database. *2 A German company selling in France. *3 Owners of the Caravell brand. *4 German mail order company. *5 Sells under the brand Foron. 5.3 DEFINITION OF SHORT-TERM STANDARDS 5.3.1 CONVENTIONAL MODELS Once the reference line of a given appliance category has been established a minimum energy standard line can be generated which is a simple factor of the energy reference line. Minimum efficiency standards have the effect of prohibiting the least efficient models from the market, such that as the standard line is strengthened a greater proportion of existing models are excluded until eventually all existing models would fail to meet the standard. There is no exact science which can be applied to establish an optimum standard line, rather the choice is founded on consideration of the following: ù the environmental benefits of reduced energy demand, ù the benefit of reduced running costs, ù the technical feasibility of meeting the standards, ù the consequences of standards for the refrigeration market, ù the consequences of standards for the manufacturers. The GEA approached this problem by defining a range of energy savings targets and working backwards to evaluate the energy standards line which the targets implied. The energy savings targets they considered where a -10% standard and a -15% standard. The savings refer to the average energy consumption of the stock of new refrigerator models for each individual category. To evaluate the -10% and -15% standards lines the GEA adopted the following algorithm. For each refrigerator category: 1 plot the energy consumption for the stock of new refrigerator models against adjusted volume for all the models in the database, 2 regress a straight line to this data which defines the equation of the energy reference line, 3 identify the least energy efficient appliance (that with the highest energy efficiency index), 4 replace it with a hypothetical unit of incrementally greater efficiency, such that the total number of appliances remains the same, 5 calculate the energy consumption of the stock of new and hypothetical refrigerator models and compare it with the energy consumption of the original stock, 6 repeat steps 3 to 5 until the target energy savings have been attained. The most subjective aspect of this process is determining how manufacturers would actually respond to any given standards line and therefore how the distribution of model energy efficiencies would differ before and after standards are introduced. The GEA process assumes that the total number of models available would not decline, which is reasonable given the American experience of standards described in the Executive Summary. For the model replacement, step 4, the GEA replaced the least efficient model by one of equal efficiency to the average of the cluster of models within a narrow band of adjusted volume about the least efficient models adjusted volume. It is possible with this approach that the savings postulated by a given standards line would not be realised in practice, because it presumes manufacturers do more than the bare minimum to satisfy the standard. This is of greater concern in the case of models which are made to conform to the standard by a simple design modification, e.g. by replacing the compressor with a more efficient type, than for those models which are replaced by a wholly new design. In many such cases the temptation will be for manufacturers to effect some change to the design which does the bare minimum to satisfy the standard. However, provided the deterrents against breaches of the standard are strong enough manufacturers are more likely to replace inefficient models with models which exceed the standard by at least a 10% margin. This is because the assembly and component tolerances are such that any individual model may exceed the energy standard if the design energy consumption is within this margin. The -10% and -15% energy standards lines and the energy reference line are shown graphically in Figures 5.1 - 5.7. The equations which describe them are shown in Table 5.4. TABLE 5.4: THE GEA PROPOSED SHORT-TERM STANDARD LINES *1 Code Maximum Energy Maximum Energy Maximum Energy Consumption Consumption Consumption Allowed Allowed (kWh/yr) Current Allowed (kWh/yr) Average Standard (kWh/yr) -15% Standard Line Line -10% Standard Line R1 Vadj*0.643+191 Vadj*0.599+178 Vadj*0.557+166 R2 Vadj*0.450+245 Vadj*0.437+238 Vadj*0.402+219 R3 Vadj*0.657+235 Vadj*0.616+221 Vadj*0.573+205 R4 Vadj*0.777+303 Vadj*0.778+303 Vadj*0.697+271 R5 & R6Vadj*0.233+245 Vadj*0.225+237 Vadj*0.207+218 F1 Vadj*0.446+181 Vadj*0.519+211 Vadj*0.480+195 F2 Vadj*0.472+286 Vadj*0.478+289 Vadj*0.433+262 *1 All adjusted volumes are in litres. The current average standard line is the energy reference line. For some of the categories setting the standard to match the energy reference line gives a greater energy saving than 10%. The anticipated energy savings from setting the standard to match the reference line are shown in Table 5.5. TABLE 5.5: ANTICIPATED ENERGY SAVINGS FOR THE ENERGY EFFICIENCY STANDARD SET AT THE ENERGY REFERENCE LINE Code R1 R2 R3 R4 R5&6 F1 F2 Energy 7.3% 8.31% 6.6% 10.1% 8.6% 18.0% 10.6% Savings 5.3.2 NO FROST MODELS No-frost models are likely to consume more electricity than manual defrost refrigerators. The GEA investigated the behaviour of no-frost models of fridge-freezers compared to the manual defrost models (category R4) and found that they had statistically indistinguishable volumes but significantly higher energy consumptions. At the time of writing their report (GEA 1993) the GEA recommended that a simple adjustment factor, F, be applied to the energy reference line as: nofrost EAV = EAV F [5.1] where F = 1.35 for no frost models and 1.0 for other models. The same factor would be applied to the energy standards line, thus to calculate the maximum energy consumption allowed for a no-frost fridge-freezer under the -10% standard, the equation for the R4 -10% standard in Table 5.4 would be solved for the given adjusted volume and multiplied by 1.35. However, the derivation of the adjustment factor F was based on the behaviour of a small number of no-frost models. Furthermore, the eco-labelling study (ENEA 1993) pointed out that it would make more sense if the correction factor was applied on a weighted basis according to the adjusted volumes of the no-frost compartments compared to the other compartments. This notion allows for the possibility that not all the compartments in the refrigerator are designed as no-frost compartments. Since this time a wider database of no-frost models has been examined and it has been agreed between researchers from the eco-labelling study, the GEA and those working on the refrigerator energy labelling scheme that a revised no-frost factor of F' = 1.2 should be applied to the adjusted volume of the no-frost compartments. This is approximately equivalent to applying a factor F = 1.1 to the energy consumption, for a model where all compartments are no-frost (Lebot 1993). This new factor is accepted for all the minimum efficiency standards calculations. 5.4 IMPACT OF SHORT TERM ENERGY EFFICIENCY STANDARDS ON NATIONAL REFRIGERATION MARKETS The GEA examined the effect of imposing the -10% and -15% standards on each of the national refrigeration markets in their database. The results are summarised in Tables 5.6 to 5.12. TABLE 5.6: PERCENTAGE OF R1 (FRIDGES WITH 1 STAR FFC) ON MARKET WHICH SATISFY THE SHORT TERM STANDARD Short Term Standards Line National Number Reference Line -10% Line -15% Line Market of Models (-7.3%) Germany 35 57.1% 48.6% 20.0% Denmark 4 50.0% 50.0% 25.0% Spain 2 50.05% 50.0% 0.0% France 41 39.0% 39.0% 0.0% Italy 11 36.4% 36.4% 27.3% The Nether'ds 16 68.0% 62.5% 18.8% Portugal 0 - - - UK 0 - - - TOTAL 109 49.5% 45.9% 12.8% TABLE 5.7: PERCENTAGE OF R2 (FRIDGES WITH 2 STAR FFC) ON MARKET WHICH SATISFY THE SHORT TERM STANDARD Short Term Standards Line National Number Reference Line -10% Line -15% Line Market of Models (-8.3%) Germany 10 20.0% 20.0% 20.0% Denmark 10 80.0% 80.0% 50.0% Spain 8 37.5% 37.5% 37.5% France 30 10.0% 10.0% 0.0% Italy 21 95.2% 57.1% 28.6% The Neth'ds 20 55.0% 45.0% 15.0% Portugal 6 50.0% 50.0% 33.3% UK 10 40.0% 40.0% 40.0% TOTAL 115 47.0% 38.3% 21.7% TABLE 5.8: PERCENTAGE OF R3 (FRIDGES WITH 3 STAR FFC) ON MARKET WHICH SATISFY THE SHORT TERM STANDARD Short Term Standards Line National Number Reference Line -10% Line -15% Line Market of Models (-6.6%) Germany 157 57.1% 48.6% 20.0% Denmark 40 65.0% 42.5% 20.0% Spain 2 100% 50.0% 0.0% France 26 38.5% 15.4% 0.0% Italy 10 10.0% 10.0% 0.0% The Neth'ds 128 58.6% 28.9% 9.4% Portugal 1 0.0% 0.0% 0.0% UK 10 80.0% 70.0% 30.0% TOTAL 374 61.5% 29.4% 11.0% TABLE 5.9: PERCENTAGE OF R4 (FRIDGE-FREEZERS) ON MARKET WHICH SATISFY THE SHORT TERM STANDARD Short Term Standards Line National Number Reference Line -10% Line -15% Line Market of Models (-7.3%) Germany 35 57.1% 48.6% 20.0% Denmark 4 50.0% 50.0% 25.0% Spain 2 50.0% 50.0% 0.0% France 41 39.0% 39.0% 0.0% Italy 11 36.4% 36.4% 27.3% The Neth'ds 16 68.0% 62.5% 18.8% Portugal 0 - - - UK 0 - - - TOTAL 109 49.5% 45.9% 12.8% TABLE 5.10: PERCENTAGE OF R5 & R6 (FRIDGES WITH CHILLERS AND FRIDGES WITHOUT FFC) ON MARKET WHICH SATISFY THE SHORT TERM STANDARD Short Term Standards Line National Number Reference Line -10% Line -15% Line Market of Models (8.6%) Germany 168 60.7% 43.5% 19.0% Denmark 81 70.4% 50.6% 22.2% Spain 20 50.0% 50.0% 40.0% France 56 17.9% 10.7% 3.6% Italy 36 44.4% 27.8% 0.0% The Neth'ds 190 38.4% 32.1% 13.2% Portugal 3 33.3% 33.3% 0.0% UK 0 - - - TOTAL 554 48.6% 36.1% 15.3% TABLE 5.11: PERCENTAGE OF F1 (CHEST FREEZERS) ON MARKET WHICH SATISFY THE SHORT TERM STANDARD Short Term Standards Line National Number Reference Line -10% Line -15% Line Market of Models (-18.0%) Germany 119 60.5% 83.2% 74.8% Denmark 80 61.3% 82.5% 68.8% Spain 25 16.0% 84.0% 44.0% France 151 38.4% 63.6% 47.0% Italy 71 16.9% 46.5% 21.1% The Neth'ds 165 47.9% 77.0% 55.2% Portugal 7 0.0% 14.3% 0.0% UK 9 0.0% 22.2% 22.2% TOTAL 627 43.7% 71.0% 53.3% TABLE 5.12: PERCENTAGE OF F2 (UPRIGHT FREEZERS) ON MARKET WHICH SATISFY THE SHORT TERM STANDARD Short Term Standards Line National Number Reference Line -10% Line -15% Line Market of Models (-10.8%) Germany 253 62.1% 63.2% 39.5% Denmark 79 43.0% 48.1% 16.5% Spain 24 62.5% 70.8% 37.5% France 187 44.4% 46.0% 20.9% Italy 55 25.5% 41.8% 3.6% The Neth'ds 225 48.4% 53.3% 25.8% Portugal 25 12.0% 12.0% 8.0% UK 17 35.3% 35.3% 29.4% TOTAL 865 48.7% 52.4% 26.4% These statistics are all based on the GEA database, which comprises a mixture of years data between 1990 and 1992 depending on the national market. The percentage of models surviving the stated minimum efficiency standards will not be static with time because, even in the absence of standards, the average model efficiency is almost certain to improve, see Section 9.2. Thus, the fraction of 1993 models satisfying the quoted standards is likely to be higher than those mentioned in Tables 5.6 to 5.12. The effect of standards on each national market for all categories combined is shown in Figure 5.43. 6: LONGER-TERM ENERGY STANDARDS 6.1 METHODOLOGY Most of the material in this section is a restatement of the GEA long-term energy efficiency standards analysis, however, new and complimentary results are provided for vacuum panel insulation options which were not fully investigated in the GEA study. Limited resources have confined this extra analysis to two product categories: fridges with no frozen food compartment (R6) and upright freezers (F2). The speed with which manufacturers can adapt their product range is a primary constraint effecting the introduction of minimum energy efficiency standards. For this reason and because technological development is not static, seperate standards need to be defined for the short- and longer-term. The statistical approach used for short-term minimum efficiency standards in Section 5 is most appropriate when standards are to be introduced within up to two years of notification. For longer notification periods manufacturers have enough time to produce wholly different models if necessary. Thus, it makes more sense to base longer-term minimum efficiency standards on a proper technical and economic analysis of the options for improving the efficiency of refrigerator designs, called the "engineering analysis". The GEA performed an engineering analysis for each refrigerator category in their study. The methodology followed was: 1 identify a base-case model, i.e. establish the theoretical physical characteristics of the "average" model in the category including its energy consumption, 2 define a set of energy efficiency improvement design options, 3 simulate the energy consumption of the base-case model after each design modification using the SPADE computer program, 4 rank the design options in the order of the most cost-effective means of saving energy, 5 take the base-case with the most cost effective design modification to be the new `base-case' and repeat steps 2 to 5 until a complete life cycle cost curve is produced, 6 set the standard at the point of lowest (or least) life cycle cost (defined below). This is essentially the same approach used to define longer-term minimum energy efficiency standards for refrigerators in the United States. Setting the standard to coincide with the lowest life cycle cost is an example of a "no regrets" policy. It produces real improvements in energy efficiency but also ensures that the consumer saves money, because savings in reduced running costs outweigh any increased model purchasing costs. Providing volume sales do not drop, manufacturers should also increase profits through the additional mark-up associated with the design changes. 6.2 CALCULATING COSTS To establish the cost of each of the considered design options the GEA originally approached Danish manufacturers to supply estimates. However, the values that they produced were challenged by the German manufacturers association which resulted in the GEA values being increased by approximately 20%. The values included the manufacturer's, distributer's and retailler's profit markup (i.e. 250%) which means that if the total purchasing cost of the refrigerator increases because of the design changes considered, the profits increase proportionally. The cost of electricity was assumed to be 0.13 ECU/kWh, which was the average marginal cost of electricity in Europe including taxes in November 1992. Simple payback periods (SPPs) were computed for each design option by dividing the increase in purchase cost by the associated saving in annual running cost, to give a first estimate of the most cost effective design option. The life cycle cost is computed by adding the refrigerator purchase cost to its discounted lifetime running costs. If the running costs, RC, are considered to be constant over the refrigerators lifespan, then the life cycle cost, LLC, is given: LCC = PP + PWF*RC [6.1] where PP is the purchase price and the present worth factor, PWF, is: PWF = (1- 1/(1+r)N)/r [6.2] N is the average operational lifespan of the refrigerator and r is the consumer discount rate, assumed to be 5%. The payback period, PBP, measures the number of years it takes to recover the extra consumer investment in increased efficiency through lower running costs: PBP = (ln(RC/PP) - ln(RC/PP - r))/ln(1+r) [6.3] ["" = delta] Providing the payback period is less than the lifespan of the refrigerator then the life cycle cost of the base-case with the extra efficiency measures is less than the life cycle cost of the base-case. 6.3 THE GEA DESIGN OPTIONS The design options considered in the GEA analysis were limited to technology that was fully proven and commercially available at the time the analysis was conducted (1992-93). Design Options Considered by the GEA: 1. Increased door insulation a) applied to the outside (increase in depth), b) applied to the inside (loss of volume), 2. Increased cabinet insulation a) with increase of appliance height (length, for chest freezers), b) with constant exterior appliance dimensions (loss of volume), 3. Increased evaporator surface area, 4. Increased condensor surface area, 5. Increased evaporator heat capacity, 6. Increased condensor heat capacity, 7. Improved compressor efficiency, 8. Decreased door leakage. For the extra insulation options, 1 & 2, the thickness was increased in steps of 15mm from the base-case values, but was limited to a maximum of 30mm extra insulation for all categories except chest freezers (F1). For chest freezers the door thickness was allowed to increase by up to 45mm and the cabinet insulation thickness was not limited. It was assumed that all insulation was PUR foam with zero ODP and a conductivity of 0.024W/mK. For the 1b & 2b options, the internal volume was ascribed a monetary value giving a penalty cost of 1 ECU per litre decreased net volume for all refrigerator compartments accept chest freezers which were valued at 0.6 ECU/litre. If possible, the evaporator and condenser surface areas, options 3 & 4, were doubled. The surface area of evaporators on fridges with frozen food compartments were not increased because they are already fully sized. Similarly, the surface area of chest freezer condensers were not altered because it would require a fundamental redesign of both the condensor and the freezer. The heat capacity of the evaporator and condenser, options 5 & 6, can be increased by using thicker walled metal piping. In both options the heat capacity was doubled. The compressor technology the GEA considered in option 7 has an EER of 6.0 and is thought to improve upon typical compressor energy efficiency by 12%. This is not quite as good as the Sun Power Inc. design of Section 4.2.1, but is still very efficient. However, the Sun Power compressor maintains a constant efficiency with cooling load and can match it's pumping capacity to the instantaneous load, which has the potential to save up to 25% more electricity. 6.4 Description of the GEA Base-case Models CATEGORY R1: FRIDGE WITH 1 STAR FROZEN FOOD COMPARTMENT Adjusted volume: 151 litres Fridge volume: 129 litres Frozen food compartment: 14 litres (VRef/VFFC=9/1) Energy consumption: 274 kWh/yr Dimensions: 92'55'60 cm3 Insulation: 30mm in top, bottom and sides 30mm in back side 50mm in outfacing sides of the frozen food compartment 20mm in the door Evaporator: Roll-bond aluminum 1mm thick, forms bottom, backside and topside in the inside of the frozen food compartment. A prolonging of the roll-bond-evaporator is making out for the refrigerator compartment evaporator. Evaporator size in the refrigerator compartment is 30'44 cm2. Condensor: Steel pipe with fins or threads H'B: 52'55 cm2 Compressor: Danfloss TL4F or similar CATEGORY R2: FRIDGE WITH 2 STAR FROZEN FOOD COMPARTMENT Adjusted volume: 204 litres Fridge volume: 169 litres Frozen food compartment: 19 litres (VRef/VFFC=9/1) Energy consumption: 335 kWh/yr Dimensions: 110'55'60 cm3 Insulation: 30mm in top, bottom and sides 30mm in back side 50mm in outfacing sides of the frozen food compartment 20mm in the door Evaporator: Roll-bond aluminum 1mm thick, forms bottom, backside and topside in the inside of the frozen food compartment. A prolong of the roll-bond-evaporator is making out for the refrigerator compartment evaporator. Evaporator size in the refrigerator compartment is 25'44 cm2. Condensor: Steel pipe with fins or threads H'B: 70'55 cm2 Compressor: Danfloss TL4F or similar CATEGORY R3: FRIDGE WITH 3 STAR FROZEN FOOD COMPARTMENT Adjusted volume: 192 litres Fridge volume: 155 litres Frozen food compartment: 17 litres (VRef/VFFC=9/1) Energy consumption: 367 kWh/yr Dimensions: 102'55'60 cm3 Insulation: 30mm in top, bottom and sides 30mm in back side 50mm in outfacing sides of the frozen food compartment 20mm in the door Evaporator: Roll-bond aluminum 1mm thick, forms bottom, backside and topside in the inside of the frozen food compartment. A prolonging of the roll-bond-evaporator is making out for the refrigerator compartment evaporator. Evaporator size in the refrigerator compartment is 10'40 cm2. Condensor: Steel pipe with fins or threads H'B: 70'55 cm2 Compressor: Danfloss TL4F or similar CATEGORY R4: FRIDGE-FREEZER WITH 4 STAR FROZEN FOOD COMPARTMENT Adjusted volume: 355 litres Fridge volume: 171 litres Freezer volume: 86 litres (VRef/VFre=2/1) Energy consumption: 591 kWh/yr Dimensions: 136'60'60 cm3 Insulation in fridge: 30mm in top, bottom, sides and backside 20mm in door Insulation in freezer: 50mm in top, bottom, sides and backside 50mm in door Evaporator: Fridge compartment: Roll-bond aluminum 1mm thick, placed inside the fridge compartment. H'B: 19'44 cm2. Freezer compartment: The evaporator is made from bent copper - or steel pipes over the drawers (3 drawers) with threads welded on the upper and lower side in approx. 0.5 cm's distance. Total length approx. 3.6m2. Condensor: Steel pipe with fins or threads H'B: 95'60 cm2 Compressor: Danfloss NL6F or similar CATEGORY R5: FRIDGE WITH CHILLER COMPARTMENT Adjusted volume: 185 litres Fridge volume: 106 litres Chiller compartment: 106 litres (VRef=VChiller) Energy consumption: 294 kWh/yr Dimensions: 115'55'60 cm3 Insulation: 30mm in top, bottom, sides and backside 20mm in door Evaporator: Roll-bond aluminum 1mm thick, placed inside the fridge compartment (21'44 cm2) and extending into the chiller compartment. Condensor: Steel pipe with fins or threads H'B: 74'55 cm2 Compressor: Danfloss TL3F or similar CATEGORY R6: FRIDGE WITHOUT FROZEN FOOD COMPARTMENT Adjusted volume: 179 litres Fridge volume: 179 litres (VRef=VAdj) Energy consumption: 301 kWh/yr Dimensions: 102'55'60 cm3 Insulation: 30mm in top, bottom, sides and back side 20mm in the door Evaporator: Roll-bond aluminum 1mm thick, placed inside the cabinet. H'B:36'44 cm2. Condensor: Steel pipe with fins or threads H'B: 67'55 cm2 Compressor: Danfloss TL3F or similar CATEGORY F1: CHEST FREEZER, 4 STAR Adjusted volume: 636 litres Freezer volume: 296 litres (VAdj=2.15'VFre) Energy consumption: 462 kWh/yr Dimensions: 85'104'65 cm3 Insulation: 60mm in top, bottom and sides. Evaporator: Integrated evaporator made of pipes on inner cabinet of sheets of aluminum. Total pipe length approximately 18m. Condensor: Integrated evaporator made of steel pipes on outer cabinet of sheets of steel. Total pipe length approximately 14m. Compressor: Danfloss NL6F or similar CATEGORY F2: UPRIGHT FREEZER, 4 STAR Adjusted volume: 358 litres Freezer volume: 167 litres (VAdj=2.15'VFre) Energy consumption: 440 kWh/yr Dimensions: 105'60'60 cm3 Insulation: 50mm in top, bottom, sides, door and back side. Evaporator: The evaporator is made from bent copper -or steel pipes over the drawers (3 drawers) with threads welded on the upper and lower side in approx. 0.5cm's distance. Total length approx. 6.5m. Condensor: Steel pipe with fins or threads. H'B:64'55cm2. Compressor: Danfloss NL6F or similar 6.5 RESULTS OF THE ENGINEERING ANALYSIS The GEA life cycle cost analysis results are shown for each category of refrigerator in Tables 6.1 to 6.8. These results are also illustrated graphically as life cycle cost curves in Figures 6.1 to 6.8. The tables give the purchase price, the annual energy consumption, the life cycle running cost, the total life cycle cost and the pay back period for the base case model with successive design modifications. The design modifications are made in order of the most cost effective energy saving measure. The emboldened design option corresponds to the point of least life cycle cost, which is shown by the minimum of the cost curve in the corresponding figure. These results are summarised in Table 6.9. Table 6.9: Energy and Cost Performance of GEA Least Life Cycle Cost Designs for Each Refrigerator Category. Category Basecase Minimum LCC Energy saving Payback Energy Period Usage PBP kWh/yr Years Energy Efficien Usage cy kWh/yr Index EEFF kWh/yr % ---------------------------------------------------------------- R1 274 141 0.49 133 49 3.1 R2 335 166 0.49 169 50 3.0 R3 367 227 0.63 140 38 3.8 R4 591 322 0.55 269 46 4.0 R5 294 149 0.52 145 49 3.9 R6 301 135 0.47 166 55 3.7 F1 462 259 0.56 203 44 4.3 F2 440 247 0.54 193 44 3.6 It can be seen from Table 6.9 that the GEA least life cycle cost designs are between 38% and 55% more efficient than the base-case designs. The higher efficiency designs would pay for their additional discounted purchase cost through lower running expenses within 4.3 years. To increase confidence in these findings the GEA performed electricity price and design options cost sensitivity analyses. Each cost curve was recomputed assuming European average marginal electricity prices of ECU 0.10/kWh and ECU 0.16/kWh and also for ñ20% design options costs. Naturally, assuming higher electricty prices and/or lower design options costs will reduce the pay back period while opposite assumptions would increase it. However, the energy consumption at the point of the least life cycle cost (the cost curve minimum) was remarkably insensitive to these variations, indicating the stability of the analysis to cost assumptions. Conversely, this result illustrates the importance of considering all the possible energy saving design options for a least life cycle cost appraisal. 6.6 THE GEA RECOMMENDATATION FOR LONG-TERM MINIMUM EFFICIENCY STANDARD LINES According to the GEA their life cycle cost analysis shows that: 1) It is technically feasible to design and produce refrigerators consuming significantly less energy than today's models. Energy efficiency standards therefore could, with a sufficient lead time (around 3 - 4 years) be set at levels considerably lower then the recommended short term standards with no technical problems. 2) The introduction of energy efficiency standards are cost effective to consumers since, according to the technical analysis, standards will produce economic savings over the lifetime of the appliance. If standards are set not lower than LCC minimums, the consumers will likely have payback periods on their investment of less than 3 or 4 years.' The GEA argued that the minimum of the life cycle cost curve should be set as the target for the long-term minimum efficiency standards. They derived the long-term minimum efficiency standard line equations from the category-specific energy reference lines by applying a fixed energy savings percentage across the whole range of adjusted volumes. The resulting standards equations are shown in Table 6.10. TABLE 6.10: THE GEA PROPOSED LONG-TERM STANDARD LINES*1 Category Maximum Energy Maximum Energy Consumption Consumption Allowed Allowed (kWh/year) (kWh/year) Current Average Long Term Standard Line Standard Line R1 Vadj*0.643+191 Vadj*0.331+98 R2 Vadj*0.450+245 Vadj*0.223+122 R3 Vadj*0.657+235 Vadj*0.406+146 R4 Vadj*0.777+303 Vadj*0.423+165 R5 & R6 Vadj*0.233+245 Vadj*0.105+110 F1 Vadj*0.446+181 Vadj*0.250+102 F2 Vadj*0.472+286 Vadj*0.265+160 *1 All adjusted volumes are in litres. The current average standard line is the energy reference line. 6.7 SOCIETAL COSTS It can be argued that it would make more sense to use the least societal cost rather than the least life cycle cost to define the energy efficiency target. The societal cost approach seeks to include the cost of environmental damages from all forms of pollution associated with the refrigerator energy use into the assessment. These extra pollution costs are known as environmental externality adders and cover: human mortality and morbidity, crop damages, forestry damages, biodiversity issues, building damages, noise pollution, visibility quality, water pollution, land pollution and global warming effects. Environmentality externality adders are difficult to estimate and a wide range of values are quoted in the literature. A quick back-of-the-envelope type analysis using adder figures from CSERGE (1992) and the average EU fuel-mix figures from the GEA indicated that the societal cost could be computed assuming an electricity price of ECU 0.17/kWh. This is only slightly higher than the high electricity cost sensitivity test investigated by the GEA, which indicates that using a least societal cost approach is unlikely to make a significant difference to the conclusions. 6.8 Least Life Cycle Costs for Vacuum Insulation Panel Designs Following the spectacular energy savings from using VIPs which were demonstrated in Section 4.2.4, Greenpeace contracted the research agency, TNO, to produce a complete life cycle cost curve for the R6 category (fridge with no frozen food compartment) but including VIPs as a design option. The VIPs were assumed to have a conductivity of 0.0029W/mK (R-50), and to cost $1.50 per square foot per inch thickness to manufacture and assemble within the refrigerator, with costs proportional to the VIP thickness. The vacuum panels covered the whole cooled space excluding gaps for the intrusion of fixtures and the door seal gasket. The cost of the PUR foam in the base-case model were not subtracted from the VIP model costs, because it was assumed that foam is used for structural support. The results are shown in Table 6.11. TABLE 6.11: LIFE CYCLE COST AND PAY BACK PERIODS OF COMBINED OPTIONS. CATEGORY R6 (VIPS AT $1.50/SQ-FT/INCH) Design Options Purch Energy Life Life Pay combined in ase consum cycle Cycle Back order of Price ption operat Cost Period increasing simple ECU kWh/yr ing ECU (years) payback. costs Category R6 ECU 0 Basecase model 485.0 301.0 347 832 1 = VIP insulation 501.6 114.5 132 634 0.7 0.5" 2 = 1 + Decreased door leakage 502.8 106.8 123 626 0.7 3 = 2 + 0.25" more 511.1 76.7 88 599 0.9 VIP insulation 4 = 3 + 0.25" more 519.4 61.5 71 590 1.1 VIP insulation 5 = 4 + 0.25" more 527.7 52.4 60 588 1.3 VIP insulation 6 = 5 + 0.25" more 536.0 46.4 54 590 1.5 VIP insulation 7 = 6 + improved 546.8 40.9 47 594 1.8 compressor 8 = 7 + 0.25" more 555.1 37.1 43 598 2.0 VIP insulation 9 = 8 + 0.25" more 563.4 34.3 40 603 2.3 VIP insulation 10= 9 + doubled evap. 573.1 32.3 37 610 2.5 capacity Under these assumptions, the least life cycle cost occurs for the fifth design option, with 1.25" thick vacuum paneling (32mm) and a pay back period of only 1.3 years. This is for an energy consumption of 52kWh/year giving an 83% saving compared to the base-case!. It should be stressed that this remarkable result is obtained without making the most favourable compressor technology assumptions (the GEA compressor cost and performance are used). If we assume that compressors are available using the Sun Power design specifications (Section 4.2.1) then it might be possible to save 25% of the energy used by the base-case without extra cost. The VIP costs used in this analysis may be considered to be at the lower end of the range. Consequently, the analysis was repeated for very much more pessimistic price assumptions, with VIPs costing $3.75 per square foot per inch (a 2.5 fold increase). The results are shown in Table 6.12. TABLE 6.12: LIFE CYCLE COST AND PAY BACK PERIODS OF COMBINED OPTIONS. CATEGORY R6 (VIPS AT $3.75/SQ-FT/INCH) Design Options Purch Energy Life Life Pay combined in ase consum cycle Cycle Back order of Price ption operat Cost Period increasing simple ECU kWh/yr ing ECU (years) payback. costs Category R6 ECU 0 Basecase model 485.0 301.0 347 832 1 = VIP insulation 526.5 114.5 132 658 1.7 0.5" 2 = 1 + Decreased 527.7 106.8 123 651 1.7 door leakage 3 = 2 + 0.25" more 538.5 76.7 88 627 1.8 VIP insulation 4 = 3 + 0.25" more 559.2 61.5 71 630 2.4 VIP insulation 5 = 4 + 0.25" more 580.0 52.4 60 640 2.9 IP insulation 6 = 5 + 0.25" more 600.7 46.4 54 654 3.5 VIP insulation 7 = 6 + improved 611.5 40.9 47 659 3.8 compressor 8 = 7 + 0.25" more 632.5 37.1 43 675 4.3 VIP insulation 9 = 8 + 0.25" more 653.0 34.3 40 693 4.8 VIP insulation 10= 9 + doubled evap. 662.8 32.3 37 700 5.1 capacity Even under these less favourable assumptions the Least Life Cycle Cost design uses 75% less energy than the base-case and has a pay back period of only 1.8 years. Both sets of results are plotted with the original GEA R6 life cycle cost curve in Figure 6.9.[FIGURE 6.9 NOT AVAILABLE] 7: GREENPEACE PROPOSALS FOR MINIMUM EFFICIENCY STANDARDS The analysis presented in the preceding text has clearly demonstrated the technical and economic desirability of introducing firm minimum efficiency standards for refrigerators without delay. After giving serious consideration to any genuine practical problems which manufacturers might incur, the following standards program is recommended. On January 1st 1994 notification of the following standards should be given: ù the introduction of the -15% standard, defined in Table 5.4, by January 1st 1995, ù the introduction of tougher minimum energy efficiency standards, defined in Table 7.1, for January 1st 1997. TABLE 7.1: GREENPEACE'S PROPOSED LONG-TERM MINIMUM ENERGY EFFICIENCY STANDARD LINES *1 Category Maximum Energy Maximum Energy Consumption Allowed Consumption Allowed (kWh/year) (kWh/year) GEA Long-term Standard Greenpeace Long-term Line Standard Line R1 Vadj*0.331+98 Vadj*0.354+105 R2 Vadj*0.223+122 Vadj*0.248+135 R3 Vadj*0.406+146 Vadj*0.447+159 R4 Vadj*0.423+165 Vadj*0.466+182 R5 & R6 Vadj*0.105+110 Vadj*0.128+134 F1 Vadj*0.250+102 Vadj*0.278+112 F2 Vadj*0.265+160 Vadj*0.293+177 *1 All adjusted volumes are in litres. In addition, the following action plan would be announced: ù The establishment of a permanent technical committee whose task it is to review the least-life cycle cost analysis in the light of new technical and cost information. This review will produce recommendations for new longer-term standards to be announced by January 1st 1997 and to commence on January 1st 2002 at the latest. Standards should be reviewed, although not necessarily changed, every three years thereafter. ù The formation of a standards registration and enforcement agency with full powers, see Section 8. The permanent technical committee responsible for standards should use clear accounting principles in setting the standards. As a bare minimum these should aim to set the standards no higher than is implied by the Least Life Cycle Cost of the appliance. Preferably the standards would be based upon the Least Societal Cost which would incorporate an assessment of the value of preventing the emission of pollutants. The technical committee should be made up of a cross-national, politically independent, research team, such as the GEA, whose findings would be binding. Standards should not be assessed by non-technical personnel in the Commission. The standards proposed in Table 7.1 are intended to have the same effect as the `long-term standards' proposed by the GEA and defined in Table 6.10. The date for introduction has been advanced from 1999 to 1997 through consideration of the successful US experience wherein all models on the market were replaced within a three year period without loss of model choice, appreciable increase in price, or any evidence of decreased manufacturer profits. In spite of this, the Greenpeace standards lines defined in Table 7.1 are 10% weaker than the GEA long-term standards lines. The GEA standards lines were set to exactly equal the targets given by the least life cycle cost analysis. However, because of component and assembly tolerances, it is anticipated that manufacturers would have to design models to be about 10% more efficient than the standards lines to be sure of complying with them. Provided the standards enforcement procedure is adequate, it seems appropriate to set the standards lines 10% above the energy efficiency target given by the least life cycle cost. 8: STANDARDS ENFORCEMENT 8.1 ENFORCEMENT ADMINISTRATION To be fully effective the enforcement administration of the minimum efficiency standards should have two components: a) initial certification of each generic models compliance with the standards, b) random inspection at point of sale to ensure that models on the market comply with the initial certification. At present manufacturers produce their own energy consumption statistics measured according to the standard EN153. The European Commission's draft proposal is that manufacturers should be allowed to test the performance of their own models for initial certification and that there should be no inspection at the point of sale. Consequently, the success of the standards would seem to be reliant upon rival manufacturers testing each others models for compliance with the standard. The danger with this lowest cost option is that testing may be ad-hoc and that manufacturers may form alliances to cover-up non-compliance. This contention is supported by the recorded discrepancy between the manufacturers energy consumption measurements and independent test results, shown in Section 2.3. To ensure that standards are properly adhered to Greenpeace proposes the following: ù The Commission should establish an independent agency to conduct both certification and random point-of-sale testing. ù Three models should be supplied by the manufacturer for initial certification with the energy rating being the average recorded under EN153. ù Should a point-of-sale test indicate that a model exceeds the standard by more than 10% a further 9 models should be tested. If the average value of all ten models is found to exceed the standard by 5% the model should be withdrawn from sale pending appeal. ù The agency should have the power to fine a manufacturer if the manufacturer is found to be guilty of deliberately trying to deceive the agency during the certification or point-of-sale testing programmes. 8.2 COST OF ENFORCEMENT ADMINISTRATION The argument put in favour of manufacturer self-policing of the minimum efficiency standards is that it is the lowest cost option at a time when expenditure by the Commission is a sensitive political issue. Nonetheless, self-policing by manufacturers does involve some extra costs which would eventually be passed on to the consumer. The life cycle cost analysis of Section 6 has shown that consumers stand to save considerable amounts of money through the introduction of meaningful energy efficiency standards. It is therefore suggested that the standards administration and enforcement scheme proposed in Section 8.1 should be financed through a small levy applied to all new refrigerators. This levy would also cover the cost of the standards technical review committee proposed in Section 7. The levy should not be viewed as an additional tax because it's effect will be to save consumers money. Rather, the levy constitutes an investment with a high rate of return. It may reasonably be argued that a failure to introduce meaningful standards and properly enforce them is akin to perpetuating an indirect historical energy tax. 9: IMPACT OF ENERGY EFFICIENCY STANDARDS In this section an analysis is presented of the impact of minimum energy efficiency standards on electricity savings and air pollution. 9.1 METHODOLOGY To evaluate the energy savings potential of the Greenpeace standards proposals the same model has been used as in the GEA analysis, the GEA `Energy Savings Model', ESM. This model forecasts the changing quantity and age distribution of the stock of fridges, freezers and fridge-freezers owned by households in each nation of the European Union until 2015. The energy consumption of the refrigerator stock is then forecast for whatever standards scenario is assumed. To compute the pollution emissions associated with the electricity consumption of the national refrigerator stocks a new emissions model has been formulated. This is similar to the GEA Environmental Impact Assessment model produced by Mark Hinnels (Appendix 7 of the GEA study), but differs because it is assumed that refrigerators have a relatively flat daily load profile and therefore use predominantly base-load electricity. Consequently, the appropriate fuel-mix to use for emissions calculations is the base-load fuel mix. The base-load fuel mix is derived from the average fuel mix by assuming that the base-load demand is 70% of the average demand and that plant is used for base-load generation in the following order of preference: 1) Nuclear 2) Coal 3) Gas/Renewables/Oil/Other It was assumed that all electricity imports are generated by nuclear plant. Although this is a simplification, by far the largest electricity exporter in the European Union is France, who has an excess of nuclear capacity. 9.2 THE BASE-CASE SCENARIO AND HISTORICAL EFFICIENCY IMPROVEMENTS Since 1970 the historical improvement in energy efficiency of refrigeration units has been approximately 2.5% per annum. This result is illustrated in Figure 9.1, which shows the recorded change in average energy consumption of Swiss refrigerators (BEW 1993) compared to the energy consumption which they would have had were the efficiency to have improved by 2.5% per annum. The trend in Swiss refrigerator energy efficiency is likely to be representative of Europe as a whole, because the Swiss market is supplied by the same manufacturers as the rest of Europe. If Swiss data is converted into annual efficiency improvements, see Figure 9.2, then it is clear that the rate of annual efficiency improvement is increasing, not slowing down as was assumed by the GEA. This is an important result because it is necessary to have confidence in the forecast gains resulting from implementing minimum energy efficiency standards and therefore to have a reasonable base-case (no standards) forecast. The Swiss data shows that it is not reasonable to assume a decreasing rate of annual energy efficiency improvement for the base-case scenario, if anything, the assumed rate should increase. Consequently, a 2.5% energy efficiency improvement per annum base-case scenario is considered to be more in keeping with the long-term mean. The GEA assumed (but did not justify) that energy efficiency would only increase at 1% per annum from 1992 to 2015. It would appear that the GEA base-case should only be viewed as a low efficiency, worst-case scenario. 9.3 RESULTS The full forecasted results from 1990 to 2015 are given in the Appendix, which shows data by country and for the European Union as a whole for the following parameters: ù fridge electricity consumption ù freezer electricity consumption ù fridge-freezer electricity consumption ù all refrigerator electricity consumption ù CO2 emissions from refrigerator electricity use ù CO2 equivalent (i.e. all greenhouse gases) emissions from refrigerator electricity use ù SO2 emissions from refrigerator electricity use ù NOx emissions from refrigerator electricity use ù total cost of refrigerator electricity paid by the consumer ù the amount of baseload generation capacity required to supply demand from refrigerators This data is given for the GEA base-case scenario, the -2.5% base-case scenario, the EU -10% scenario, the Greenpeace A scenario and the Greenpeace B scenario. Additionally the savings from Greenpeace A and B scenarios relative to the GEA base-case are shown. Cumulative data is supplied for the EU as a whole, showing the cumulative output/savings up to the given date. The collective EU data is shown for some of these parameters in Figures 9.3 to 9.7 (`no energy standards' refers to the GEA base-case in these figures). The ineffectiveness of the draft EU proposal is fully apparent from this analysis as the EU standards scenario is shown to probably give no real benefit. However, the huge potential for meaningful gains is indicated through the Greenpeace standards scenarios. It is believed that benefits akin to those shown for the Greenpeace B scenario would result from full implementation of the standards setting methodology adopted in the United States as applied in the European context. 10: THE CONTRIBUTION OF HFCS TO THE GLOBAL WARMING COMMITMENT OF A TYPICAL REFRIGERATOR In this section a simple analysis is presented which illustrates the importance of avoiding the use of HFCs as refrigerant and foam blowing substitutes to CFCs. HFCs are not ozone depleting but they are very potent greenhouse gases as seen through the Global Warming Potentials (GWP) of a variety of common HFCs, shown in Table 10.1. TABLE 10.1 THE GLOBAL WARMING POTENTIAL OF HFCS *1 Direct Effect for Time Horizons of Gas Lifetime Sign of (years) 20 years 100 years 500 years "Indirect" Effect HFC-125 40.5 5200 3400 1200 none HFC-134a 15.6 3100 1200 400 none HFC-143a 64.2 4700 3800 1600 none HFC-152a 1.8 530 150 49 none * 1 Source (IPCC 1992) The GWP is the relative radiative forcing of a greenhouse gas compared to CO2. It is dependent upon the time horizon considered; thus, a given quantity of HFC-134a has a global warming effect equivalent to 3100 times as much CO2 if the effect is assessed over 20 years from release to the atmosphere and 1200 times a much CO2 if the effect is assessed over 100 years. This information can be used to assess how significant a contribution HFCs might make to the global warming impact of a refrigerator over its lifespan if they are used to blow the insulant foam and also used in the refrigerant. For simplicity, a typical three star refrigerator using the most commonly adopted CFC substitute, HFC-134a, is considered. The following assumptions were made: ù there is no significant increase in energy consumption for a fundamentally identical design of refrigerator if HFCs are used in place of CFCs or HCFCs in both the refrigerant and the foam (GEA 1993), ù the standard three star refrigerator prior to controls on CFC content used 140g of CFC-12 as a refrigerant and 202g CFC-11 as the foam blowing agent (Hinnels 1993), ù HFCs will replace the CFCs on a one to one basis, i.e. the total HFC content of the refrigerator is 342g (GEA 1993), ù the Global Warming Potential of a kilogram of HFCs is 1500 times a kilogram of CO2 (IPCC 1992), ù the average 1992 three star refrigerator consumes 367 kWh of electricity per year (GEA 1993), ù The European average emissions of CO2 to generate 1 kWh = 0.549kg (see Section 9), ù When other Greenhouse Gases (GHGs) are included the European average emission of GHGs to generate 1 kWh = 0.646 kg CO2 Equivalent (Section 9), ù A typical refrigerator lifespan is 15 years (Hinnels 1993). ù All HFCs are ultimately released to the atmosphere. From these assumptions it is possible to compute the direct and indirect TEWI (Total Equivalent Warming Impact) for the average 1992 three star model from: Indirect Emissions = 367 * 15 * 0.646 = 3391.5 kg CO2 Equivalent Direct Emissions = 0.342 * 1500 = 513 kg CO2 Equivalent These results indicate that for an average three star refrigerator using HFCs in the foam and the refrigerant that the HFCs contribute 15% of the total equivalent warming impact. As refrigerators become more energy efficient the relative contribution of the direct (HFC) component is likely to increase, especially if the greater energy efficiency is a result of increasing the insulation thickness. This effect is shown in Figure 10.1 where the change in the direct and indirect contribution is indicated as a function of the energy consumption for hypothetical designs of an otherwise typical three star refrigerator. As the energy consumption falls in response to increasing the insulation thickness the indirect component decreases but the direct component increases, until the direct component gives 34% of the refrigerator TEWI for a model consuming 61% of the energy of the base-case refrigerator. These results indicate that the use of HFCs as substitute gases for CFCs is incompatible with the goal of minimising the environmental impact of refrigerators, the more so if one considers the ready availability of pentane and butane substitutes, which have almost no impact. REFERENCES ANIE (1993) Ecolabel, Statistical Information About Fridge and Freezers, cited in Study for the Attribution of a European Ecolabel for Refrigerators and Freezers Draft Report, ENEA (Italian Agency for Energy, Environment and New Technologies), June. Appliance Magazine, Dana Chase Publications Inc., 1110 Jorie Boulevard, CS 9019, Oak Brook, Illinois 60522-9019. Arch, B (1993,1994) Personal communication with Owens-Corning representative. Benson, D.K. & Potter, T.F. (1992) "The effect of new priorities and new materials on residential refrigerator design" National Renewable Energy Laboratory, American Council for an Energy Efficient Economy (ACEEE), 2.13-2.21. BEW (1993) Data base complied by Bundesamt fuer Energiewirtschaft and cited in (GEA 1993) CA (1992) U.K. Consumers Association, internal document analysing refrigerator energy efficiency. Colley, T. (1993) Personal communication. Researcher testing vacuum panels at Oak Ridge National Laboratory for U.S. D.O.E. CSERGE (1992) The Social Cost of Fuel Cycles Report to the UK Dept. of Trade and Industry by The Centre for Social and Economic Research on the Global Environment (CSERGE), September, HMSO, London. EIU (1989) Marketing in Europe 322, The Economic Intelligence Unit, September, pp47. EIU (1990) Marketing in Europe 337, The Economic Intelligence Unit, December, pp56. EIU (1991) Marketing in Europe 340, The Economic Intelligence Unit, March, pp46. ENEA (1993) Study for the Attribution of a European Ecolabel for Refrigerators and Freezers Draft Report, ENEA (Italian Agency for Energy, Environment and New Technologies), June. Euromonitor (1993) European Major Electrical Appliance Companies Euromonitor report, May. Fines, H.A. (1992) Vacuum Panel and Thick Wall Foam Insulation for Refrigerators: cost Estimates for Manufacturing and Installation US Environmental Protection Agency, EPA/430/R-92/110, October. Forbes Pearson, S. (1991) The Replacement of CFC Refrigerants report for Star Refrigeration Ltd, Glasgow, UK. G&A (1992) Accessions for Great Britain to 31st March 1992 G & A Marketing Services Ltd. GEA (Group for Efficient Appliances) (1993) Study on Energy Efficiency Standards for Domestic Refrigeration Appliances Final report for DG-XVII of The Commission of the European Communities, The GEA comprise researchers from: Agence de l'Environnement et de la Maitrise de l'Energie (France), Danish Energy Agency, Novem (The Netherlands), TNO (The Netherlands), dk-TEKNIK (Denmark), LASH-ENTEPE (France), CCE (Portugal). Hinnels, M (1993) 'Parallels between energy efficiency in domestic appliances and environmental life cycle analysis of products', paper for Institute of Advanced Studies, Manchester Metropolitan University. IPCC (1992) The Intergovernmental Panel on Climate Change Scientific Assessment, 1992 Update report. ISO (1988) Draft Revision of ISO 8187 (8187.3): Performance of Household Refrigerating Appliances - Refrigerator-freezers International Standards Organisation, Paris. Jeffs, M. (1993) Personal communication. Representative of ICI Belgium working on encapsulated open cell vacuum panels. December 1993. Larder, D. (1993) Personal communication. Testing engineer for U.K. Consumers Association. Lebot, B. (1992) Personal communication. Researcher with the GEA working for ADEME in France. Lebot B. (1993) Personal Communication with GEA researcher, December. McMahon, J.E. (1992) "Quantifying the Benefits and Costs of U.S. Appliance Energy Performance Standards", Lawrence Berkely Laboratory, Energy and Environment Division, Energy Analysis Program. McNaught C. (1993) "Domestic refrigeration controllers", CADDET Energy Efficiency Newsletter (Centre for the Analysis and Dissemination of Demonstrated Energy Technologies supported by the IEA and OECD) 3 p13-4. Michel P. (1992) "Analysis of Two Door Fridge/Freezers Number of Compressors and Number of Thermostats" Internal paper for GEA, Laboratoire Sciences de l'Habitat, France. Missendon, J.F. & James, R.W. (1992) "The use of propane in domestic refrigerators", International Journal of Refrigeration 15 p95-100. Norg†rd, J.S. (1989) Electricity: Efficient End-Use and New Generation Technologies, and Their Planning Implications Lund University Press, Denmark. Pederson, P.B. (1992) Engineering Analysis Concerning Energy Efficiency Standards for Refrigerators and Freezers in Denmark Preliminary study, report to the Danish Energy Agency and Norwegian Water Resources and Energy Administration, Technical University of Denmark, Lund, February. Perruzo, R. (1993) `Refrigerator/freezers energy consumption measurements', letter from Zanussi representative. Remich, N.C (1993) "50% more energy efficiency with CFC-Free Insulation" Appliance Manufacturer January, 53-4. Reuter, R. & Sextl, G. (1993a) "Vacuum insulation panels", presented at: 1993 Non-Fluorocarbon Insulation, Refrigeration and Air-Conditioning Technology Workshop, September 28th, Wiesbaden, Germany. Reuter, R. & Sextl, G. (1993b) "Life expectancy of household appliances using vacuum insulation panels", presented at: International CFC & Halon Alternatives Conference, October 20-22, Washington D.C. Simpson, D. (1993) "Compressors" Appliance May, 51-57. Turiel, I., Berman, D., Chan, P., Chan, T., Koomey, J., Lebot, B., Levine, M., McMahon, J., Rosenquist, G, & Stoft, S. (1992) "U.S. Residential Appliance Energy Efficiency: Present Status and Future Directions" Building Equipment and Appliances June, 1213-33. Turiel, I. (1993) Personal communication with researchers in the Lawrence Berkeley Laboratory Waide, P.A. (1993) Mandatory Energy Efficiency Standards for Domestic Refrigeration Units in the European Community Report for Energy Policy Research Unit of Greenpeace International.