The costs and benefits of introducing standards and labels for electrical appliances in Pacific Island countries

Transcription

1 The costs and benefits of introducing standards and labels for electrical appliances in Pacific Island countries

2 The costs and benefits of introducing standards and labels for electrical appliances in Pacific Island countries Prepared for the Secretariat of the Pacific Community Supported by Department of Climate Change and Energy Eiciency and AusAID September 2011

3 Copyright Secretariat of the Pacific Community (SPC) 2012 All rights for commercial / for profit reproduction or translation, in any form, reserved. SPC authorises the partial reproduction or translation of this material for scientific, educational or research purposes, provided that SPC and the source document are properly acknowledged. Permission to reproduce the document and/or translate in whole, in any form, whether for commercial / for profit or non-profit purposes, must be requested in writing. Original SPC artwork may not be altered or separately published without permission. Original text: English Secretariat of the Pacific Community Cataloguing-in-publication data for electrical appliances in Pacific Island countries / prepared by George Wilkenfeld and Associates Pty Ltd for the Secretariat of the Pacific Community 1. Electric apparatus and appliances Standards Oceania. 2. Energy labeling Oceania. 3. Energy conservation Oceania. I. Title II. George Wilkenfeld and Associates III. Secretariat of the Pacific Community AACR2 ISBN: Design and layout: Muriel Borderie SPC Publications Section Prepared for publication and printed at the Secretariat of the Pacific Community, Noumea, New Caledonia, 2012 ii

4 Acknowledgements Prepared by George Wilkenfeld and Associates Pty Ltd, Energy Policy and Planning Consultants, PO Box 934, Newtown NSW 2042, Sydney, Australia. Tel (+61 2) The author is grateful for the comments and suggestions of several reviewers, including Mr Peter Johnston (Technical Consultant), Mr Anthony Maxwell (Asian Development Bank), Mr Brian Dawson (Secretariat of the Pacific Community) and Mr Tim Farrell (Department of Climate Change and Energy Eiciency Australia). iii

5 Foreword The United Nations has declared 2012 as the International Year of Sustainable Energy for All. It is an opportunity for the global community to give special attention to the central role sustainable energy plays in our socio-economic and sustainable development. Access to aordable and sustainable energy transforms lives by enabling people to produce goods and services more eiciently and become more productive members of their communities. It creates jobs and generates much needed income and lifts people out of poverty. It transforms the lives of families and communities. It supports better health, education and communication services through, for example, powering water pumps, community health clinics, schools and cell phones. It empowers women by reducing the time and drudgery involved in collecting and using traditional energy sources and creates a cleaner environment with minimal contribution to climate change. Achieving Sustainable Energy for All is a challenge given that everyone is entitled to have access to energy sources that are reliable, aordable, clean and safe. Energy eiciency and renewable energy are the two pillars of sustainable energy. It is estimated that 7 million out of the 10 million people living in Pacific Island countries and territories still do not have access to any form of electricity. Therefore, the region has largely concentrated on generating extra electricity. In its role as the lead agency coordinating the implementation of the new Framework for Action on Energy Security in the Pacific, SPC with the support of key partners, including the Australian Government Department of Climate Change and Energy Eiciency, is now also focusing attention on more eicient use of the electricity already available. Minimum Energy Performance Standards and Energy Rating Labels for electrical appliances are two of the most cost-eective regulatory interventions that can be adopted as part of the region s eorts towards providing sustainable energy for all. This report is the first that examines the costs, benefits and opportunities for energy standards and labelling in Pacific Island countries and territories. It puts forward information for policy makers and indicates the potential benefits that can be realised by governments, businesses and consumers alike. Dr Jimmie Rodgers Director-General Secretariat of the Pacific Community iv

6 Contents Acknowledgements... iii Foreword...iv Summary...1 Recommendations Introduction Energy labelling and MEPS... 7 Energy labelling...7 MEPS...9 MEPS and product quality Appliance energy programs in Pacific Island countries Studies in the mid-1990s...11 Studies since This study Electricity use in Pacific Island countries Sources of data Access to electricity supply Electricity use by sector Household appliance ownership and electricity use Commercial and government sector electricity use Electricity generation costs Energy eiciency measures Potential of energy eiciency by main end use Refrigeration Air conditioning Lighting...32 Summary Projected Energy Saving Impacts Residential Commercial/government Combined eects Benefits and costs Projected benefits Costs and benefit/cost ratios Costs of electricity avoided or generated Greenhouse gas abatement costs Conclusions and recommendations...52 Conclusions Recommendations References v

7 Figures Figure S1. Projected electricity use , by sector...2 Figure S2. Projected electricity saving from energy eiciency measures, by PIC...3 Figure 1. Estimated annual electricity use by main sector, PICs Figure 2. Estimated total electricity use per capita, PICs Figure 3. Estimated household electricity use per person living in electrified home, PICs Figure 4. End use shares of household electricity use, by PICs...20 Figure 5. Average electricity use per electrified household in 2010, by PIC...21 Figure 6. Total electricity use in the residential sector in 2010, by PIC...22 Figure 7. Approximate electricity use in the commercial/government sector in 2010, by PIC...23 Figure 8. Approximate end use share of electricity used in residential and commercial/ government sectors combined, by PIC...23 Figure 9. Crude oil price (TAPIS Blend), Figure 10. Energy-eiciency trends, refrigerators and freezers, Figure 11. Average electricity use per unit sold new, Figure 12. Short-term market response and longer-term stock response to refrigerator MEPS in PICs...31 Figure 13. Average annual energy consumption per residential appliance or energy use, BAU and with energy eiciency measures...31 Figure 14. Preliminary estimates of potential for commercial/government end uses...33 Figure 15. Projected electricity use , by sector...37 Figure 16. Projected electricity savings by end use...37 Figure 17. Projected electricity savings from energy eiciency measures, by PIC...38 Figure 18. Annual electricity cost savings per capita (all sectors) medium oil price (USD 100/bbl)...42 Figure 19. Annual electricity cost savings per household (residential) medium oil price (USD 100/bbl)...43 Figure 20. Value of annual electricity supply costs savings and potential CDM CO2 credits, all PICs medium oil price...43 Figure 21. Electricity supply cost curve (undiscounted) medium oil price...47 Figure 22. Electricity supply cost curve (7% discount rate) medium oil price...47 Figure 23. Emission reduction resource cost curve (undiscounted) medium oil price...50 Figure 24. Emissions reduction resource cost curve (7% discount rate) medium oil price...51 vi

8 Tables Table S1. Summary of modelling outcomes, all 14 PICs...1 Table S2. Summary of value of projected electricity savings, all PICs (USD million)...2 Table 1. Products covered by mandatory energy labelling and MEPS in Australia and New Zealand...9 Table 2. Data sources used...13 Table 3. Estimated population and electrification rates, PICs, Table 4. Estimated appliance ownership rates in electrified households, PICs...18 Table 5. Range of appliance ownership and usage attributes, 14 PICs...19 Table 6. Comparative residential electricity taris, PICs Table 7. Summary of typical generation cost assumptions...25 Table 8. Factors determining the marginal cost of diesel generation...27 Table 9. Projected energy use indices under BAU and energy eiciency assumptions in PICs...33 Table 10. Projected growth rates in electricity use, BAU and energy eiciency scenarios...35 Table 11. Summary of modelling outcomes, all 14 PICs...35 Table 12. Projected electricity savings , by end use...36 Table 13. Summary of value of projected electricity savings, All PICs (USD million)...39 Table 14. Value of projected electricity-related savings , by PIC undiscounted lower oil price (USD 75/bbl)...39 Table 15. Net present value of discounted electricity-related savings , by PIC lower oil price (USD 75/bbl)...40 Table 16. Value of projected electricity-related savings , by PIC undiscounted medium oil price (USD 100/bbl)...40 Table 17. Net present value of discounted electricity-related savings , by PIC medium oil price (USD 100/bbl)...41 Table 18. Value of projected electricity-related savings , by PIC undiscounted higher oil price (USD 125/bbl)...41 Table 19. Net present value of discounted electricity-related savings , by PIC higher oil price (USD 125/bbl)...42 Table 20. Indicative cost of renewable electricity generation options...46 Table 21. Energy eiciency and renewable generation options ordered by cost...46 Table 22. Net cost of emissions savings from energy eiciency programs...48 Table 23. Net cost of emissions savings from renewable energy generation...49 Table 24. Emissions reduction options ordered by cost...50 vii

9 viii

10 Summary This is a study of the potential for Minimum Energy Performance Standards (MEPS) and energy labelling programs to increase the eiciency of electricity use in Pacific Island countries (PICs). It covers electricity production and use in all the member countries of the Pacific Islands Forum, with the exception of Australia and New Zealand. The study is largely based on existing data and research, including a detailed 2004 study of the potential for MEPS and labelling applied to refrigerators and freezers in Fiji, PIC census data and reports of the Pacific Islands Renewable Energy Project (PIREP 2005). As the quality of data for PICs is highly variable, the findings of this study with regard to any individual PIC should be considered preliminary and should be verified with further research. The cost-benefit modelling in this study concentrates on the potential for electricity savings through greater eiciency in refrigeration, lighting and air conditioning and estimates the potential benefits (in monetary units, energy units and greenhouse gas emissions units) to individual PICs. It was assumed that each PIC would apply the same MEPS and energy labelling regimes as currently apply in Australia and New Zealand (ANZ). This would be the least costly and most eicient option because many appliances sold in the PICs are manufactured in or supplied from ANZ, and it would be possible for PIC governments to tie in to the existing administrative structure of the ANZ program. The northern PICs, which use 110V supply, are obviously an exception, but they account for about 6% of the electrified households covered in this study. The main economic value of MEPS and energy labelling in the PICs would be a reduction in the projected demand for imported diesel fuel for electricity generation and a reduction in diesel maintenance and capital costs. Although all PICs have some renewable energy resources, renewable generation is highly capitalintensive and variable on a daily and seasonal basis, so it is projected that diesel will continue to meet the majority of the growth in electricity demand in the medium term. In many PICs domestic electricity users benefit from both consumer class cost subsidies (e.g. from commercial and government users) and geographical subsidies in that taris are often uniform irrespective of the actual cost of supply. This means that the value to PIC economies of increasing the eiciency of household electricity use is generally higher than taris alone indicate. The marginal costs of diesel generation in each PIC will depend on a range of factors. The most critical and most likely to fluctuate are crude oil prices and the local currency to USD exchange rates (PNG is the only PIC with its own oil and gas supplies). The study models costs and benefits under three levels of Tapis crude oil price: a low price of USD 75 per barrel, a medium price of USD 100 and a high price of USD 125. Given the range of generation and distribution eiciencies, the medium oil price equates to diesel generation costs of USD 0.24 to 0.37 per kwh actually delivered to users (all cost estimates are in US dollars rather than local currency). 1

11 Table S1 and Figure S1 summarise the main outcomes of the modelling, for all 14 PICs combined. Over the period , it is projected that the population of the PICs will increase by nearly 25% (based on ADB projections) and the average electrification rate across the PICs will increase from about 22% to 29%. In combination, these two factors are expected to raise the number of electrified households by nearly 74%. For residential, commercial and government sectors combined, energy eiciency measures would mean that projected electricity use in 2025 would be about 2672 GWh instead of 3031 GWh, a saving of about 12%, or 359 GWh per year. Table S1. Summary of modelling outcomes, all 14 PICs Change Population 8,584,598 10,710, % Persons/HH % Households 1,682,556 2,193, % Electrification rate 22% 29% 33.2% Electrified HH 367, , % Total res GWh (BAU) 691 1, % Total res GWh (EE) 691 1, % GWh saved/yr % Res kwh/hh (BAU) 1,882 2, % Res kwh/hh (EE) 1,882 2, % Res kwh/hh saved % Total comm/govt GWh (BAU) % Total comm/govt GWh (EE) % GWh saved/yr % Comm kwh/cap (BAU) % Comm kwh/cap (EE) % Total GWh demand (BAU) 1,606 3, % Total GWh with measures (EE) 1,606 2, % Total GWh saved % BAU: Business as usual; EE: Energy eiciency Lighting and air conditioner energy saving measures each account for about 37% of the total savings, and refrigeration accounts for about 27%. This ignores any increase in the energy eiciency of other sectors, about which so little is known that no estimates are possible. Because of its large population and growing electrification rate, it is estimated that PNG would account for about 38% of the electricity savings. Fiji would account for about 29% and the other 12 PICs for a third (Figure S2). Again, these are preliminary estimates based on incomplete data. 2

12 The monetary benefits of these energy savings are estimated at between USD 582 million and USD 895 million (undiscounted) over the period 2010 to 2025, depending on the oil price (Table S2). The avoidance of future fuel costs accounts for between 81% and 88% of the benefit, depending on the oil price. Maintenance and interest cost savings account for between 10% and 15% of the value. Emissions avoided over the period 2011 to 2025 would total about 2,230 kt CO 2- equivalent (CO 2 -e). If the value of this could be realised as Clean Development Mechanism (CDM) credits, at USD 10 per tonne CO 2 -e, it would make up between 2% and 4% of the total value of savings. Table S2. Summary of value of projected electricity savings, all PICs (USD million) Lower oil price (USD 75/bbl) Medium oil price (USD 100/bbl) Higher oil price (USD 125/bbl) Undiscounted, real dollars USD 582 M USD 739 M USD 895 M Net present value (7% discount rate) USD 262 M USD 332 M USD 403 M GWh electricity used per year Combined (BAU) Combined (EE) Res (BAU) Res (EE) Comm/Govt (BAU) Comm/Govt (EE) Figure S1. Projected electricity use , by sector 3

13 400 GWh per year saved by EE programs Vanuatu Tuvalu Tonga Solomon Islands Samoa Papua New Guinea Palau Niue Nauru Marshall Islands Kiribati Fiji FS Micronesia Cook Islands Figure S2. Projected electricity saving from energy eiciency measures, by PIC The projected monetary savings per electrified household range from USD 274 per year in Nauru to USD 61 per year in Solomon Islands, with an average for all PICs of USD 95 per year per electrified household. 1 The main costs to PIC governments of developing and establishing a MEPS and energy labelling regime are: Ongoing administrative costs (i.e. the salaries and on-costs of government, electricity authority or NGO sta responsible for developing legislation, administering the program, providing a product registration service, checking compliance in the field etc.). Costs of establishing databases of refrigerators, freezers, air conditioners and other products that may be subject to MEPS or energy labelling. These costs could be minimised by making use of the existing electronic register that supports energy labelling and MEPS in Australia and New Zealand. Costs of publicising the establishment of mandatory energy labelling and MEPS to retailers and the general public. Costs of carrying out compliance checks on products sold. It will be very important, especially in the early years, to have a reasonable level of checktesting of products to ensure suppliers and the general public have confidence in the program. Again, it should be possible to make use of laboratories in nearby countries. The main costs to the public are those associated with the purchase of more eicient appliances and products than would be the case without MEPS or energy labelling. Some of these costs will come from the fact that energy eicient products will probably cost more than less eicient products, although studies in countries with a history of energy eiciency programs suggests that this is not necessarily the case. 1 These estimates are based on published electricity taris, or for PICs where there are no meters, on the estimated costs of generation. 4

14 It would be necessary to undertake detailed studies of the appliance and equipment market in each PIC to estimate the likely costs of imposing higher energy eiciency requirements on imported equipment. The only study to date, covering the refrigerator and freezer market in Fiji in 2004, indicated that the costs would be between a quarter and a tenth of the value of the benefits (giving benefit/cost ratios between about 4:1 and 10:1). If a benefit/cost ratio of 5:1 were realised across all the PICs, it would mean that by 2025 the average electrified PIC household would be paying USD 12 to 55 more for appliances and lamps each year, but realising between USD 61 and 274 in annual energy cost savings (depending on the oil price). The projected economic benefits of MEPS greatly exceed the projected costs, both for product buyers and for PICs as a whole. Indeed, given that the marginal generation fuel is usually diesel, and that the marginal electricity tari appears to be well below the cost of supply in most PICs, each kwh avoided has a higher value to the economy than to the individual householder. In eect, it is far less costly for PICs to import more eicient refrigerators, air conditioners and lights than to import diesel fuel. It is possible to directly compare the cost of avoiding electricity use through greater energy eiciency with the costs of supplying the equivalent amount of energy by renewable energy generation, including wind turbines, micro-hydro and photovoltaics (PV). It is estimated that, given the mix of large and small electricity grids, the weighted average cost of delivering a kwh of diesel-generated electricity to end users in the PICs is about USD 0.31 (taking losses into account). The energy eiciency measures covered in this report could save electricity at an average cost of USD 0.05 per kwh. This is about USD 0.26 per kwh less than the cost of diesel generation and about USD 0.60 per kwh less than the cost of PV generation. Energy eiciency programs also have the added benefit of avoiding or reducing greenhouse gas emissions. As the value of diesel fuel and other generating costs avoided would be greater than costs of energy eiciency programs, there would be net savings of USD 573 million (undiscounted), or USD 276 million at a discount rate of 7%. This means that emissions would be saved at a negative cost of USD 257 per tonne CO 2 -e avoided (undiscounted) or USD 124 per tonne avoided (at a discount rate of 7%). The cost of achieving similar emission savings though a mix of wind, micro-hydro and PV generation is estimated at a positive cost of USD +244 per tonne CO 2 -e avoided (undiscounted) or USD +173 per tonne avoided (at a discount rate of 7%). This is because grid-connected renewable generation costs in the main PIC population centres are usually higher than diesel generation costs. In more remote areas, renewable generation may be cheaper than diesel, but expensive battery storage or diesel backup would be necessary to give a level of electricity service that is comparable to the grid. In eect, saving greenhouse gas emissions in the PICs using renewable generation alone would cost 9 to 11 times more than doing so through programs that increase the eiciency of electricity use. However, these cost-benefit estimates are only preliminary, and significantly more research would be required to make specific estimates for each country. 5

15 Recommendations The following recommendations are made: 1. A small number of Pacific Island countries (PICs) should be selected for implementation of standards and labelling for refrigeration, air conditioners and lighting. The PICs that can provide on-the-ground program support should be prioritised. 2. The standards and labelling program in PICs should be aligned with other regional programs such as the Equipment Energy Eiciency (E3) program that is being delivered jointly by Australia and New Zealand. 3. The standards and labelling program in PICs should be overseen by a central organisation in the Pacific. The Secretariat of the Pacific Community (SPC), as the central agency for energy in the Pacific, may be the most appropriate agency to assist in the development of a standards and labelling in the Pacific. 4. Once a standards and labelling program has been established in a small number of PICs it could be expanded to other PICs and include a greater range of energy using equipment. 5. Improved energy consumption data in PICs would enable the prioritisation of products to ensure the program maintains high impact and cost-eectiveness. The studies in these countries should cover: a) The main end uses of electricity; b) The source of import of key appliances; c) The ownership and technical characteristics of appliances and equipment; d) The channels for appliance and equipment imports and sales; e) The energy sources and costs of generation; f) The institutional and administrative framework to support energy eiciency programs; g) The standards and labelling program in PICs also needs to develop an evaluation strategy to quantify the energy, economic and possibly environmental benefits. 6. The standards and labelling program in PICs also needs to develop an evaluation strategy to quantify the energy, economic and possibly environmental benefits. 6

16 1. Introduction 1.1 Energy labelling and MEPS Energy labelling and Minimum Energy Performance Standards (MEPS) are the most widespread policy measures taken by governments around the world to increase the energy eiciency of appliances and other electrical equipment. About 50 countries, as well as the European Union (EU), now have energy labelling or MEPS programs. 2 Energy labelling Energy labelling is a system that allows buyers to compare the energy eiciency of the products they are considering purchasing. The following types of information might be made available to buyers: 1. The energy consumption (say in kwh per year for electrical appliances) for each specific model when tested according to a given technical standard and assuming a certain pattern of usage. 2. A visual indicator of the energy eiciency of each model e.g. by means of a star rating. 3. An indicator that a product exceeds a specified level of energy eiciency, or that it possesses a certain capability e.g. the Energy Star label used for computer equipment (originated in the USA, but now used around the world). The form in which the information is made available might also vary. It may be on a tag or sticker attached to the product itself so that buyers can see it when they go to a store or showroom where the product is displayed. The same information could also be included in product brochures and advertisements, so that buyers can become aware of a product s energy consumption and levels of eiciency even before they visit a showroom. This information may be on packaging so that customers buying from warehouses can use this information. If the energy information on all models is collected into a printed guide or a register accessible on the Internet, it is easier for buyers to get an idea of how the energy eiciency of a model they are interested in compares with others, even if the others are not displayed in the same showroom. It can also create demand for the more eicient models since buyers interested in energy can identify and seek them out more easily. Governments might be interested in promoting energy eiciency for a range of strategic reasons, but energy labelling will only work if product suppliers and buyers consider it in their interest to prefer more energyeicient products. For buyers, the incentive is largely financial: the expectation that a more energy-eicient product will be cheaper to run. It might cost a little more to buy than a less eicient product (although this is not always the case) but this is still worthwhile if the running costs are low enough

17 To work out the running cost of any energy labelled product, the buyer needs to multiply the energy consumption information on the label (which might be in kwh per year) by the appropriate energy tari (e.g. cents per kwh). 3 If energy labelling has the intended eect on appliance buyers, product suppliers should respond by introducing and promoting more eicient models, and removing their less eicient ones from the market. However, the extent to which dierent manufacturers, importers, suppliers and retailers can respond will vary. Importers who have access to a range of brands and products may be able to obtain more eicient products to meet a market demand. Importers tied to a single supplier of products which are not very energy eicient may have to discount them heavily to obtain sales in a more energy conscious market, or in the extreme case withdraw from the market altogether. Labelling will increase the commercial value of a good energy rating, create a commercial penalty for a poor rating and could add to supplier costs if the products have not already been energy tested. This could enable less scrupulous suppliers to: Understate the energy consumption of their products. Produce entirely fictitious labels for products that have never been energy tested. Suppliers may also be tempted to reduce other aspects of product performance in order to get a good energy rating (e.g. by producing refrigerators that do not keep food as cold, or washing machines that do not wash clothes as well), but in Australia and New Zealand this is prevented by setting minimum performance levels. It is not possible to legitimately claim an energy rating or prove compliance with MEPS unless the product also meets those performance levels. Of course unscrupulous suppliers could as easily make false claims about a product s performance as about its energy eiciency, so it is important to have some means of checking claims. The incentive for products to meet performance and labelling requirements will be increased if there is an eective compliance and monitoring regime, backed with appropriate legislation. Energy labelling works best where there is a reasonable choice of models on the market, and they have dierent levels of energy eiciency. If there are only a few models and they are all at similar levels of eiciency (e.g. all have 3 or 4 stars on the label), then labelling will not have much scope to influence consumer choice. 3 It would be more direct to have the actual annual running cost on the label (e.g. in dollars per year), but this is diicult in practice. The same model with the same label may be distributed in many markets with dierent currencies and energy taris, so it is impossible to ensure that the running cost data on a standard label are accurate for each market. Even in the same market, taris change over time and dierent customer classes may be subject to dierent taris. 8

18 MEPS Energy labelling provides buyers with information that is consistent and reliable and enables consumers to take into account the energy costs of an appliance at the time of purchase. From this perspective, it can increase the eiciency of market operation through better information. However, energy labelling does not directly force suppliers to introduce more energy eicient products or to remove the less eicient ones from the market, so if appliance buyers are not sensitive to labels the program will have little eect. 4 MEPS, on the other hand, set a legally enforceable minimum level of energy eiciency, which ensures that the energy savings are achieved provided that suppliers comply with the regulations. Labelling and MEPS programs can and do work together. Australia and New Zealand for example, have both programs operating in parallel (Table 1). Energy labelling and MEPS are complementary in their impacts. MEPS remove the least eicient models from the market while labelling enables buyers to select the most eicient of the models remaining. The most eective combination of measures depends on how the market for a particular product operates. Labelling has limited eect for products that are not purchased in showrooms or are purchased by builders or plumbers rather than the person paying the energy bills, so MEPS is the more eective strategy. Even so, suppliers can choose to label if they wish, as some do if their products are particularly energy eicient (e.g. if they meet designated high eiciency criteria). The products for which labelling is optional are indicated in Table 1. Note that this is not the same as voluntary labelling. A supplier who opts to label must use the label specified by law and must comply with all requirements or face penalties in the event of proven non-compliance. Energy labelling and MEPS are complementary in their administrative basis as well as in their impacts. They rely on the same energy tests and the same information base on products. Once a labelling program is in place, the cost of implementing MEPS is marginal, and once MEPS are in place, the cost of implementing labelling is relatively small provided, of course, that both programs are based on the same tests and protocols. It would not be workable for one country to have a labelling program based on the Australian system, for example, and a MEPS program based on the US system. Neither MEPS nor labelling will be eective unless the rules are clear and applied equally to all product suppliers. Otherwise suppliers, retailers and customers will quickly lose confidence in the scheme. Mandatory labelling can be a powerful element in reinforcing compliance with a MEPS program. Without a label, product retailers or oicials responsible for checking compliance will find it far more diicult to verify the model number, the supplier or the claimed energy performance of a given product. Rather than reading this o the label, they will have to find the model number on the compliance plate, check it against a register and track down the importer or manufacturer. If the trail of documentation is incomplete, there is a greater chance that breaches will not be followed up. This will be known to unscrupulous suppliers, who will be much more prepared to take the risk of supplying products that fail MEPS. If there is a requirement to label as well as meet MEPS, there can be a presumption that any product that lacks a label also fails MEPS, so checking and compliance eorts can be concentrated more eiciently. 4 It can still have some initial eect if suppliers expect buyers to respond and introduce some high-eiciency models accordingly, but this eect will diminish once it become apparent that customers do not prefer the more eicient models. 9

19 Table 1. Products covered by mandatory energy labelling and MEPS in Australia and New Zealand Product or product group Measure Residential Other Household refrigerators & freezers Energy labelling 1986 P Label enhancements 2000, 2008 MEPS 1999, 2005 Electric storage water heaters (large) MEPS 1999 P Electric storage water heaters (small) MEPS 2005 P Electric storage water heaters (miscellaneous) MEPS 2005 P Clothes washers, dishwashers, clothes dryers Labelling 1986, 1990 P Label enhancements 2000 Household air conditioners Energy labelling 1986 Label enhancements 2000, 2010 MEPS Packaged air conditioners MEPS 2001, 2010 P Chillers MEPS 2009 P Close control air conditioners MEPS 2009 P Televisions Labelling 2009 P MEPS 2009 Set top boxes MEPS 2009 P External power suppliers MEPS 2009 P P Commercial refrigeration (display cabinets) MEPS 2004 P Fluorescent lamp ballasts MEPS 2003 P P Linear fluorescent lamps (tri-phosphor) MEPS 2005 P P Incandescent lamps MEPS 2009 P P Motors (3 phase) MEPS 2001, 2006 P Power supply transformers MEPS 2004 P Standby energy (range of products) One-watt target, 2013 P Swimming pool & spa equipment MEPS 2011 P Gas water heaters MEPS 2009 P Gas space heaters MEPS TBC P Gas ducted heaters MEPS TBC P Personal computers & monitors MEPS TBC P P Water heaters Greenhouse Standards 2010 P Clothes washers, dishwashers Energy impacts of WELS 2006 P P P Source: E3 (2009) MEPS and product quality MEPS legislation gives the government legal powers to prohibit the import of products that do not meet the specified levels of energy eiciency. The same powers could also be used to prohibit the import of products that fall below a specified level of performance, for example, with regard to power factor or tolerance to voltage fluctuations. It is understood that the lower quality compact fluorescent lamps (CFLs) sold in some PICs are particularly sensitive to supply voltage fluctuations, and so have short service lives. MEPS could be used to ensure that only products that meet selected quality standards are imported. 10

20 1.2 Appliance energy programs in Pacific Island countries Studies in the mid-1990s Most of the large electrical appliances sold in PICs are imported from Australia or New Zealand (ANZ). Many of these arrive with their ANZ energy labels, so these have been seen in some PICs since the early 1990s. However, as labelling has not been mandatory, retailers have been free to remove them before putting products on display. Also, even where there were labels, not all appliance buyers understood them or used them in their purchase decisions. The possibility of building a PIC energy labelling program on the ANZ label was first considered about 15 years ago. The Demand Side Management Potential study funded by the United Nations Development Programme (UNDP) in 1995 recommended energy labelling and MEPS for refrigerators, freezers and air conditioners in the 10 PICs studied (SRCI 1995). In 1996, the South Pacific Forum Secretariat commissioned a study of the scope for the systematic application of energy labelling and MEPS to increase energy eiciency in Fiji and other PICs (GWA and EES 1996). The study recommended the implementation of mandatory energy labelling and MEPS programs, based on the ANZ program, initially in Fiji, Papua New Guinea and Tonga, and eventually other PICs. The appliances recommended for labelling were refrigerators, freezers and air conditioners. MEPS were recommended for refrigerators, freezers and electric storage water heaters. MEPS and labelling were not recommended for other products, because they were not significant energy users in the PICs or because there was no ANZ MEPS regime for them, or both. Studies since 2000 The Fiji Department of Energy (FDOE) commenced work on a voluntary labelling system as early as In 2001, FDOE, with the assistance of major retailers, compiled a list of the most widely available refrigerator and freezer models with their ANZ energy label details, including kwh/yr and star rating. In 2002 and 2003, FDOE printed 40,000 brochures and distributed 90,000 flyers promoting the energy label, and also promoted the label in TV and cinema advertising, consumer newsletters and press releases. FDOE worked with a number of large retailers to: Encourage the stores to increase the proportion of floor stock carrying energy labels. Distribute the brochures listing the labelled refrigerators. Train sales sta to promote and explain the label to customers. FDOE also carried out surveys of consumer awareness of the label and the priority they give to energy eiciency, both at the beginning and at the end of the promotion campaign. In 2004 the Australian Greenhouse Oice (forerunner of the Department of Climate Change and Energy Eiciency) commissioned George Wilkenfeld and Associates (GWA) to review the impacts of the FDOE program and to assess the costs and benefits of implementing a mandatory energy labelling and MEPS regime for refrigerators, freezers and other appliances in Fiji. 11

21 The study concluded that energy labelling did not appear to be very eective in encouraging refrigerator buyers in Fiji to prefer more energy-eicient refrigerators. The fact that labelling was not mandatory at the time was only part of the reason (in fact, labelling rates were relatively high for a voluntary program because so many models are imported from New Zealand or Australia with their labels already attached). The more significant reasons for the limited impact of voluntary energy labelling of refrigerators in Fiji were: Appliance buyers appeared to be very motivated by first cost, so lower running costs had less impact on their purchase decision even if they were aware of the label. The electricity taris faced by some domestic consumers were below the costs of supply, so consumers undervalued running costs in the purchase decisions. 5 The range of models on the market was smaller than in Australia or New Zealand, so there were fewer opportunities to make a decision among two or three equally acceptable models on the basis of energy eiciency. These conditions are also common to other Pacific Island countries. These barriers will not be overcome by mandatory energy labelling alone. However, they can be addressed through mandatory MEPS, provided these can be eectively enforced. The regulatory and administrative infrastructure for implementing MEPS is almost identical to that for mandatory energy labelling, so both could be implemented at the same time, especially as the information and support infrastructure is already in place in Australia and New Zealand. Mandatory labelling would be an important and highly cost-eective means of encouraging suppliers to comply with MEPS, since tracing product suppliers and verifying performance claims would be far easier if there are labels on the products. The 2006 GWA study concluded that the economic benefits of MEPS would greatly exceed the projected costs, both for refrigerator buyers and for the Fiji economy as a whole. In eect, it was far less costly for the Fiji economy to import more eicient refrigerators (and other electrical appliances) than to import diesel fuel. The report concentrated on domestic refrigeration only, as there was not enough information to carry out an equally rigorous analysis for other products. Also, domestic refrigeration was particularly important in Fiji, accounting for over half of domestic electricity use at the time, although this may well have changed since the introduction of new types of appliances. At least two other product groups accounted for significant shares of electricity use in Fiji and are covered by MEPS in Australia and New Zealand: Air conditioners, which it was estimated in the study accounted for about 8% of household electricity and 39% of commercial sector electricity use. Commercial refrigeration and ice-making equipment, which accounted for about 11% of commercial sector electricity use. The study recommended that these should also be considered for MEPS, once the regulatory and administrative infrastructure was in place. However, it was recommended to commence with refrigerators since most suppliers were familiar with both labelling and MEPS and compliance was achievable at relatively low cost. 5 Although the Fiji Electricity Authority (FEA) generally covers its costs overall, uniform taris mean that there are often crosssubsidies between customer classes and especially between concentrated and dispersed load centres. This means that electricity users in outlying or remote areas usually do not face the full costs of supply. 12

22 Further study of electricity end use in Fiji may come to dierent conclusions. For example, it is understood that there may be particular issues with regard to the performance and quality of products such as CFLs that could be addressed through MEPS. 6 This study This study was commissioned by the Australian Department of Climate Change and Energy Eiciency (DCCEE).The objectives were to: Apply the detailed cost-benefit modelling carried out for refrigerator and freezer energy labelling and MEPS in Fiji (in GWA 2006) to other PICs. Apply the more detailed modelling of energy labelling and MEPS for lighting and air conditioning in Australia to the PICs (taking account of the best available information on how those technologies vary in the PICs). Apply the overall impact of the Australian energy labelling and MEPS program across the board (to the extent feasible) to all the PICs. The aim is to estimate the range of potential benefits (in monetary units, energy units and greenhouse gas emissions units if possible) to individual PICs. 6 However, more research is necessary to determine the importance of these products in dierent sectors or areas. About 90% of lamps in grid-connected or generator-supplied rural Fiji households in 2006 were linear fluorescents and only about 8% were incandescents, where there is scope to substitute CFLs. In households served by photovoltaic (PV) systems, however, 100% of lamps were CFLs (FDOE 2006). 13

23 2. Electricity use in Pacific Island countries 2.1 Sources of data This study covers electricity production and use in the Pacific Islands Forum member countries, with the exception of Australia and New Zealand. The use of Internet search engines was the only means of research used for this project. The main sources of information on the total electricity production of each country up to 2003 and rates of household connection to electricity supply were the country reports of the Pacific Islands Renewable Energy Project (PIREP 2004) and, where available, later reports by the local electricity supply authority. These data sources were also used to allocate electricity sales between household, commercial/hospitality/government and other sectors. Country censuses were another important source of data on electrification rates and appliance ownership by households. Not all PIC censuses collect this information. Data sources used are indicated in Table 2. Relevant development agency and consultant reports were also researched; the most valuable of these are listed in the references. As data sources reported information for dierent years, it was necessary to standardise population, electrification and electricity usage rates to Given the age and incompleteness of some of the data, it was ultimately necessary to use the author s discretion, and all estimates should be considered approximate. Table 2. Data sources used PIREP (a) Censuses (b) Electricity authority Energy plan Cook Islands X X 2005 FS Micronesia X X 2000(d) Fiji X X 2005 X 2009 Kiribati X X 2005 Marshall Islands X X Nauru X X 2002 Niue X X 2006 Palau X X 2005 Papua New Guinea X Samoa X X 2006 X 2006 Solomon Islands X X 2007 Tonga X X 2006 X (c) Tuvalu X X 2005 Vanuatu X X 2009 (a) Country reports of Pacific Islands Renewable Energy Project (PIREP 2004) accessed via pirep.htm (b) Latest available census and household income and expenditure data with material on electrification or appliance ownership, accessed via (c) Data extracted from World Bank (2010) (d) 2010 Census results not yet available 14

24 2.2 Access to electricity supply There four main categories of household with respect to appliance use: (a) Households connected to the national electricity grid. (b) (c) (d) Households with local or village diesel or micro-hydro generators. Households with access to photovoltaic supply. Households without any form of access to electricity. The PIC censuses are not always clear about the dierences between categories (a), (b) and (c). Only categories (a) and (b) are able to use the relatively high-energy appliances that are the subject of this study. Category (c) households usually have enough battery storage to support lighting and electronic equipment, but cannot operate standard mains voltage refrigerators or other appliances. In this study electrified households are those that are either grid-connected or have suicient generation of their own to operate at least an AC refrigerator, as well as lighting and electronic equipment. The estimated shares of households and number of households in this category in each PIC are summarised in Table 3. Several of the smaller PICs have electrification rates approaching 100%. However, the largest number of electrified household is in Fiji, and the second-largest number is in PNG, because even though PNG has the lowest electrification rate (10%) it also has the highest population by far. The overall rate of electrification in the PICs is estimated at about 22%, but if PNG is excluded the rate is 52%. Table 3. Estimated population and electrification rates, PICs, 2010 Population Persons/ household(a) Households Electrification rate HH with electricity HH without electricity KWh/electrified HH Cook Islands 19, ,757 97% 3, ,342 FS Micronesia 107, ,189 50% 8,542 8,648 3,882 Fiji 944, ,944 70% 132,261 56,683 1,554 Kiribati 112, ,152 51% 10,277 9,874 1,173 Marshall Islands 64, ,904 70% 9,033 3,871 6,173 Nauru (b) 14, ,804 99% 2, ,909 Niue 1, % ,660 Palau 20, ,159 99% 4, ,575 Papua New Guinea 5,940, ,188,155 10% 118,816 1,069,340 1,865 Samoa 219, ,333 93% 27,259 2,074 1,500 Solomon Islands 595, ,380 13% 14,609 97,770 1,035 Tonga 120, ,210 79% 16,669 4,541 1,206 Tuvalu 12, ,475 98% 2, ,197 Vanuatu 243, ,029 22% 10,126 35,903 3,000 All PICs 8,418, ,649,771 22% 360,841 1,288,930 1,924 Excluding PNG 2,477, ,616 52% 24, ,591 1,954 (a) Where actual data not known, average household size of 5.0 has been assumed (b) Data unreliable, according to PIREP (2004) it is possible that some electricity used in phosphate production was combined with household use in previous years and that this data conflation has not been corrected since 15

25 2.3 Electricity use by sector There are three main sectors of electricity use in the PICs: Electricity consumed by electrified households (both grid-connected and with own generation). Electricity consumed in commercial buildings (hotels, shops, oices etc.) and government buildings (schools, hospitals, oices etc), and public lighting. Electricity consumed in manufacturing, large-scale agriculture (e.g. sugar refining in Fiji) and mining (e.g. resource development in PNG). Some of these use their own generation rather than the public grid, and often supply local homes as well. Figure 1 illustrates the estimated annual energy use by the first and second of the above sectors. The scale is truncated so that electricity consumption in the smaller PICs is not swamped by manufacturing/ resources use in Fiji (190 GWh) and PNG (2213 GWh). In total, about 17% of electricity used in the PICs is for households, about 23% for commercial and government use and 60% for manufacturing or resource uses, almost all in Fiji and PNG. Figure 1. Estimated annual electricity use by main sector, PICs 2010 Source: Author estimates based on data sources in Table 2 Figure 2 illustrates total electricity use per capita, with total electricity consumption (residential plus other) divided by total population. This roughly indicates the electricity-intensity of the economy. Figure 3 focuses on household electrification rates and the intensity of household electricity use (residential sector electricity consumption energy divided by estimated population living in electrified households only). It also indicates the percentage of homes electrified in each PIC. 16