1 4c(4 4 public disclosure authorized - world bank · 2016. 8. 5. · hcfc hydrochlorofluorocarbon...

1 4c(4 4

OZONE LAYERPROTECTION

COUNTRY INCREMENTAL COSTS

Edited byKenneth King

andMohan Munasinghe

Global Environment Facility (GEF)4 andThe World Bank

ESD

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Ozone Layer ProtectionCountry Incremental Costs

Edited byKenneth King

andMohan Munasinghe

Global Environment Facility (GEF)and

The World Bank

Copyright i 1995The International Bank for Reconstructionand Development/The World Bank1818 H Street, N.W.Washington, D.C. 20433 U.S.A.

This publication may be reproduced in whole or in part and in any form for educational ornonprofit uses, without special permission from the copyright holder, providedacknowledgment of the source is made. The Secretariat of the Global Environment Facilitywould appreciate receiving a copy of any publication that uses this publication as a source.Copies may be sent to the Environment Department, The World Bank, 1818 H Street N.W.,Washington, D.C. 20433, U.S.A. and to the GEF Secretariat, G 6005, 1818 H Street, N.W.,Washington, D.C. 20433.

No use of this publication may be made for resale or other commercial purpose without theprior written permission of the copyright holders.

The designations of geographical entities in this book, and the presentation of materials, donot imply the expression of any opinion whatsoever on the part of the World Bank or the GEFor their member institutions concerning the legal status of any country, territory, or area, orof its authorities, or concerning the delimitation of its frontiers or boundaries.

The interpretations and conclusions in this report are those of the authors and do notnecessarily represent the views of the World Bank or the GEF or their member institutions.

The Global Environment Facility (GEF) is an entity that provides grants and concessionalfunds to developing countries for projects and activities that aim to protect the globalenvironment. The GEF Secretariat is functionally independent but administrativelysupported by the World Bank. It is headed by a CEO/Chairperson who reports to the GEFCouncil-its governing body The GEF's imnplementing agencies are the United NationsDevelopment Programme (UNDP), the United Nations Environment Programme (UNEP),and the World Bank. In addition the World Bank is the Trustee of the GEF Trust Fund.

ISBN 0-8213-3133-7

CONTENTS

Foreword ..................... v

Acknowledgments ......................... vii

Abbreviations and Acronyms ..................... ix

About the Contributors ..................... xi

Introduction ......................... xiii

1. Country Incremental Costs of Phasing Out Ozone-Depleting Substances .1Kenneth King and Mohan Munasinghe

2. Evaluation of Incremental Cost Methodologies for Phasing Out Ozone-DepletingSubstances: Case Studies of India, Turkey, Jordan, and Zimbabwe .25Caroline Mason

Annex 2-1-Unit Cost Assumptions .34

3. Evaluation of Incremental Cost Methodologies for Phasing Out Ozone-DepletingSubstances: Case Study of Egypt .51Charles B. Catanach

4. Evaluating the Incremental Costs of Complying with the Montreal Protocol: Discussionof a Framework Developed for Thailand .63Vikram Widge, Mark Radka, and Thomas Dillon

5. Evaluation of Incremental Cost Methodologies for Phasing Out Ozone-DepletingSubstances: Case Studies of Tunisia and Czechoslovakia .83Ulla Blatt Bendtsen

iii~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

FOREWORD

The World Bank's Environment Department promotes the development ofpractical methodologies for environmental costing and valuation. It does so inorder to assist operational divisions of the Bank and other concernedorganizations in applying these techniques. One of the methodologies pioneeredand developed in the Bank is the calculation of the incremental costs of projectseligible for financing under the Multilateral Fund of the Montreal Protocol onSubstances that Deplete the Ozone Layer. The Bank was motivated to establishthis methodology because it is an implementing agency of the Multilateral Fundof the Montreal Protocol and has in fact used that methodology as the basis forits operations. Since then, we have broadened this methodological developmenteffort in two ways: by examining the incremental costs of actions directed atother global environmental problems, and by examining the issues in calculatingincremental costs at sector and country levels.

The Global Environment Facility (GEF) provides finances to eligible countriesto phase out ozone-depleting substances, to reduce emissions of greenhousegases, to protect biodiversity, and to reduce the pollution of international waters.(In fact, GEF, which has recently been restructured and replenished, is the interimoperating entity of the financial mechanisms for the Framework Convention onClimate Change and the Convention on Biological Diversity.) These activitiesimpose added burdens on countries already struggling to mobilize financialresources for national development priorities. That is, they impose "incrementalcosts." For this reason, GEF is keen to evaluate and extend the application of thepragmatic methodologies developed in the Bank for measuring such costs. Inaddition, GEF has its own program of policy studies-the Program for MeasuringIncremental Costs for the Environment (PRINCE)-which will further develop andextend the Bank's initial work into the other focal areas of its operations.

GEF collaborates with several institutions in the framework of PRINCE. Manyof these are regional centers of excellence in developing countries whichundertake policy-related studies and act as focal points in their regions for thedissemination of training material on incremental costs. Other collaboratorsinclude various organizations involved in policy studies or research.

The papers in this collection cover the application of incremental costmethodologies to the issue of phasing out ozone-depleting substances indeveloping countries. They were first presented at the Workshop on Country-Level Incremental Costs of Phasing Out Ozone-Depleting Substances, jointlyhosted by the Environment Department and the GEF in Washington, D.C., onNovember 30, 1993. The department had commissioned the studies from thefirms that had applied the methodology in the course of preparing country

v

programs for the Multilateral Fund. The studies sought to test how well themethodology worked-not to make recommendations for operations in thecountries in question, which is the objective of the country programs themselves.The department and the GEF jointly reviewed the studies and drew conclusionsof a generic nature that will help the preparation of future country costing studiesin ozone as well as other focal areas.

We are particularly gratified by the enthusiastic response of the organizationsthat have contributed their time and effort to this evaluation and by the successfulcollaboration between the World Bank and the GEF in carrying out this work.

Mohamed T. El-Ashry Andrew SteerChairman and Chief Executive Officer Director, Environment DepartmentGlobal Environment Facility The World Bank

vi

ACKNOWLEDGMENTS

We gratefully acknowledge the invaluable contributions of the chapterauthors-Charles B. Catanach, Thomas Dillon, Ulla B. Bendtsen, CarolineMason, Mark Radka, and Vikram Widge-as well as the helpful ideas and adviceprovided by John Christensen, Claus J0rgensen, and Fraser Morrison. Thanks arealso owed to the following organizations for whom the authors work:cowIconsult, Denmark; ICF Incorporated, United States; Touche RossManagement Consultants, United Kingdom; and UNEP Collaborating Centre forEnergy and the Environment, Denmark.

Finally, we are most grateful to Rebecca Kary and Jay Dougherty of Alpha-Omega Services, Inc., and Stephanie Gerard for assistance in the editing andproduction stage, as well as to Judith Smith of Soleil Associates, who designedthe cover for the series.

vii

ABBREVIATIONS AND ACRONYMS

CBD Convention on Biological DiversityCEE Central and Eastern EuropeCFC chlorofluorocarbonsCFCCOST CFC Cost Model (developed by cowiconsult)cis Commonwealth of Independent States (former U.S.S.R.)CMEA Council for Mutual Economic Assistance, also known as

COMECON, and comprising Cuba, Mongolia, and VietnamCOSCOM Country Study Cost Model (developed by ICF Incorporated)CSFR Czech and Slovak Federal RepublicCTC carbon tetrachlorideDIW Department of Industrial Works (Thailand)FCCC Framework Convention on Climate ChangeFOB free-on-boardGEF Global Environment FacilityGHG greehouse gasGNP gross national productHCFC hydrochlorofluorocarbonHF hydrofluoric acidBFC hydrofluorocarbonIFCT Industrial Finance Corporation of ThailandLPG liquefied petroleum gasOLADE Latin American Energy OrganizationMAC mobile air conditionerMCF methyl chloroform (1,1,1 -trichloroethane)NOP National Ozone PolicyNPV net present valueO&M operations and maintenanceODP ozone-depleting potentialODS ozone-depleting substanceOORG Ozone Operations Resources GroupPRINCE Program for Measuring Incremental Costs for the EnvironmentPVC polyvinylchlorideTDRI Thailand Development Research InstituteUAC unit abasement costuc unconstrained demandUCCEE UNEP Collaborating Centre for Energy and the EnvironmentUNEP United Nations Environment Programme

it

ABOUT THE CONTRIBUTORS

Charles B. Catanach is a consultant specializing in the phaseout of ozone-depleting substances. He is based in Mullica Hill, New Jersey.

Thomas Dillon is a consultant with ICF Incorporated, Washington, D.C.

Kenneth King is Senior Environmental Specialist with the Secretariat of theGlobal Environment Facility, based in Washington, D.C.

Ulla Blatt Bendtsen is Senior Consultant with cowIconsult, Copenhagen.

Caroline Mason is Senior Consultant with Touche Ross ManagementConsultants, London.

Mohan Munasinghe is Chief of the Environmental Economics Division at theWorld Bank in Washington, D.C.

Mark Radka is a consultant with ICF Incorporated, Washington, D.C.

Vikram Widge is a consultant with ICF Incorporated, Washington, D.C.

xi

INTRODUCTION

The Vienna Convention for the Protection of the Ozone Layer 1985 was thestarting point for global cooperation to protect the stratospheric ozone layer. Itwas followed by the Montreal Protocol on Substances that Deplete the OzoneLayer in 1987, the London Amendment in 1990, and the CopenhagenAmendment in 1992. However, complying with the Montreal Protocol and itsAmendments does constrain industrial development and therefore adds to costs.These costs, the incremental costs, would be incurred by anycountry-developed, developing, or in transition-when it made special effortsthat it would not otherwise have made, in order to comply.

At the time of the London Amendment it was recognized that the incrementalcosts incurred by developing countries would place a heavy burden on theirdevelopment efforts. Therefore, the Multilateral Fund was established to channelresources from developed to developing countries to meet the agreed incrementalcosts of phasing out ozone-depleting substances (ODS). Likewise the GlobalEnvironment Facility (GEF), which also finances projects that phase out ODS,

meets each incremental cost.Both the Multilateral Fund and the GEF require a strategic framework for the

activities they finance to demonstrate that overall phaseout of ODS will beaccomplished. This framework is the Country Program, which sets out a nationalstrategy and a program of proposed activities to meet the phaseout schedules setout in the Montreal Protocol. The overall cost of the phaseout is the countryincremental cost of implementing the entire Country Program. Knowledge of thecountry incremental costs will help to prioritize and program activities and toestimate the aggregate resources required for a complete phaseout worldwide.Demonstrating the feasibility of this approach would also suggest usefulapproaches to other global environmental issues such as reducing greenhouse gasemissions.

The framework presented in Chapter 1 by King and Munasinghe wasdeveloped to estimate the country-level incremental cost. This framework hasbeen applied in several developing countries and countries in transition. Thepurpose of the Workshop on Country-Level Incremental Costs of Phasing OutOzone-Depleting Substances was to gather key analysts engaged in this work toreview the framework and its practical application. They present the results oftheir work in India, Turkey, Jordan, and Zimbabwe (Chapter 2: Mason); Egypt(Chapter 3: Catanach); Thailand (Chapter 4: Widge, Radka, and Dillon); andTunisia and Czechoslovakia (Chapter 5: Bendtsen).

xiii

These country studies all succeeded in applying the broad framework. Theirmodeling approaches differed, of course-simpler models were used when notall the stages of production, application, and consumption of ODS wererepresented in the country industrial stnicture. Full optimization was used in onlyone case, since in others the phaseout problem is highly constrained. A numberof specific modeling issues and suggestions for further developing the data andapproach also emerged from these studies and are summarized in Chapter 1.While it was not always possible to calculate the absolute level of countryincremental cost with confidence, the relative costs of alternative programschedules were quite stable. In most countries the "accelerated schedule" is eitherthe most cost-effective one or close enough to it, so that adopting it would be thepreferred risk-management strategy.

As there are many similarities between ODS phaseout and international actionon other global environmental issues (like greenhouse warming), many conceptsand modeling approaches developed for the ozone depletion issue will prove tobe useful in the analysis of those other issues. However, one of the mostimportant lessons follows from the observation that incremental cost calculationsfor ODS phaseout appear to be highly constrained. This implies that there are fewdegrees of freedom and few opportunities to explore cost-effective solutions. Toavoid this situation in dealing with other issues, it would be necessary toconstruct treaty targets, schedules, and eligibility requirements more broadly.

Kenneth KingMohan Munasinghe

xLv

1

Country Incremental Costsof Phasing Out Ozone-Depleting Substances

Kenneth King and Mohan Munasinghe

Incremental Costs

UNDERLYING ALL RATIONAL DECISIONMAKING is the conceptof the effect of each possible action. The situation expectedto result from a course of action (the with situation) iscompared to what might have been expected otherwise, inthe absence of that action (the without situation). Explicitlyor implicitly, the extra cost of the contemplated action isweighed against the added benefit that is expected.

1

2 OzoNE LAYER PROTECTION: COUNTRY INCREMENTAL COSTS

The "extra" (or incremental) cost is the Overall, we would like to minimizedifference between the costs of the with (or global incremental costs-that is, to achievealternative case) and the without (or baseline a given level of ozone-depleting substancescase). By comparing incremental costs and (ODS) phaseout at the lowest cost to theincremental benefits, the best decision may world community. To achieve this in prac-

be made to proceed. In economic cost-bene- tice, disaggregate incremental costs may befit analysis, this decision rule is expressed considered at different levels: project, sec-by requiring that incremental benefits must tor, and country. It is generally least costlyexceed incremental costs for the action to be to take action at the highest level, which is

justified. usually the country level. Alternately, inMany actions that help protect the global many cases, it is not even feasible to con-

environment incur incremental costs. For sider with or without actions at the projectexample, the choice of more expensive (but level because of the inherent linkages be-non-ozone-depleting) technologies and tween projects and sectoral activities. Ac-chemicals to provide a given level of refrig- tions to phase out ODS production cannot beeration yields a global environmental benefit considered independently of actions toin the form of protection for the strato- reduce dependency on ODS in various appli-spheric ozone layer. This global benefit has cations, because action to do one will im-not been valued monetarily but has been pose costs on the other unless the two arejudged implicitly to exceed the costs of dealt with simultaneously. Likewise, in thephasing out ozone-depleting substances power sector, it is usually not possible to(ODSs). Incremental costs are being incurred express a decision in the form of a choice

to protect global biodiversity, reduce the risk between one individual project (e.g., a wind-of climate change, and prevent the pollution mill) and another (e.g., a coal-fired powerof international waters. station). Rather, one needs to compare the

The benefits of protecting the global or respective integrated long-range powerregional environment accrue to many na- capacity expansion plans, of which eachtions, rather than only to the country that project forms a part.incurs the cost of the action. Although theaction may be justified economically from Country-Level Incremental Coststhe viewpoint of the entire global commu-

nity, it may impose what Richard Benedick In this report we are concerned withterms an "added burden" financially on a calculating the country-level incrementalgiven country. However, by allocating at costs of actions within developing countries

to protect the global environment. Specifi-least this incremental cost to the intema- cally, we are concered with the method of

tional community as a whole, the country calculation, the robustness of the calcula-undertaking the action will be left no worse tion, and the usefulness of the results foroff financially. Cost sharing is particularly determoining least incremental cost strate-important to developing countries because gies, sequencing investments, and measur-

they are unable to bear the financial burden gie "added burden." and studur-implicit in protecting the environment. It have been undertaken by many groups, suchwas in light of this fact, and to make neces- as the underake Uny groups

sary resource transfers, that financial mecha- Environment Programme, and the Unitednisms such as the Multilateral Fund of the Nations Development Programme (ODS

Montreal Protocol and the Global Environ- phaseout); the United Nations Environment

ment Facility were set up. Programme (biodiversity); and by the

United Nations Development Programme

1. Country Incremental Costs of Phasing Out Ozone-Depleting Substances 3

Collaborating Center for Energy and the gate the problem by removing its causes.Environment (UCCEE) and others on reduc- That is, we are concerned with costs in-ing the emissions of greenhouse gases. curred through phasing out the ODSs-not

Although we do not address the specific the costs that are imposed anyway by adap-requirements or procedures set out for the tation to or damage resulting from ozonecountry studies or country programs re- depletion, by international trade restrictionsquired by the various global environmental on nonsignatory or noncomplying countries,conventions-the Montreal Protocol, the or by increases in the economic cost ofFramework Convention on Climate Change producing ODS and ODS technologies for a(FCCC), and the Convention on Biological shrinking international market.Diversity (CBD)-the questions above are A similar distinction may be drawn be-implicit in all such country studies. tween the imposed costs of climate change

(adaptation costs and residual damage,Incremental Costs of Phasing Out measured against a baseline of no climate

Ozone-Depleting Substances change) and the costs of actions that reducethe risk of climate change (costs of actions

In an earlier paper (King and Munasinghe that reduce net emissions of greenhouse1991), we established a framework for the gases, measured against a baseline of cli-ozone issue and described a way to calculate mate change but without any specificallythe incremental cost to both countries and global consideration being taken into ac-firms of actions to phase out ozone-deplet- count).ing substances. That paper was the basis for Although we distinguish the incrementaldiscussions with various consulting firms costs of mitigation from those of adaptation,who subsequently undertook country-spe- we do not de-emphasize the fact that thecific calculations in selected developing global environmental benefits of mitigationcountries. are shared, nor do we imply that imposed

costs should not be compensated in someSome Important Distinctions way. Our intention is rather to draw atten-

We made several distinctions (King and tion to the fact that analysis of these twoMunasinghe, Section II) in that paper in types of costs must proceed in different

' . ~~~~~~~~ways. In this report we have chosen to dealorder to clarify what was meant by incre- with. mIthig osts becaue the toperamental costs in different settings. We re- . . tgtion dogso is oreur-viewed these distinctions in the light of gent-gien tha mitgton is aene (rsubsequent work and found that they contin- men7 .imed at phenomenons whiex aate-ued to be useful-in fact, they can be used ton ismediate and occrn the futuein other global environmental focal areas aswell. Below, we restate the principal obser- Second, we distinguish the economic

vations. issue of the incremental cost from the finan-Fatirnst,we distinguishbetweenimposed cial policy issue of sharing that incremental

Fis,w itnus ewe moe cost. In this report we are concerned withcosts and costs voluntarily accepted in order .th en issues wh ane cons'itto mitigate the global environmental prob- e ependently of fan qestons:lem. Although imposed costs could also be e incem detl of is inep tiofexamined in an incremental cost framework,this is not our specific intention here. The who will finance it. Indeed, many countries

not applying or eligible for financial assis-tance will nevertheless want to know the

those that are incurred voluntarily by a incremental economic costs of their actions.country in order to prevent or at least miti-

4 OZONE LAYER PROTECTION: COUNTRY INCREMENTAL COSTS

The economic issue is rooted in the with- Country incremental costs are economic.without dichotomy that underlies all rational The Multilateral Fund, which reimbursesdecisionmaking, not only in the global project incremental costs at the enterpriseenvironmental arena, or in the economic level, considers only the economic costs ofarena. To reach a decision, the incremental projects to be eligible and specifically ex-cost of the action (the difference between cludes transfer payments. (Implicitly, it isthe cost with and without the action) needs assumed that governments would waive theto be compared with the incremental benefit, incremental taxes and duties, that lost subsi-even if only implicitly, and whether or not dies would not be compensated, and thateither has been monetized. incentive structures would be the responsi-

This distinction between economic and bility of the governments.)financial issues is particularly important in Fourth, we distinguish project-level,the context of global environmental prob- sector-level, and country-level incrementallems because the two issues have often been costs. The project-level incremental costconfused. The economic issue is a technical arises from that part of the capital invest-one. The incremental cost defines what a ment and operations and maintenance ex-country would have to give up (trade off, penditures on the substitute plant that issacrifice) in order to meet its commitments incremental to the originally justified invest-under a global environmental convention (to ment and operations and maintenance ex-help protect the global environment)." penditures and from the economic lossWhile the solution to this equation is not associated with accelerated replacement ofalways immediately evident, finding the capital stock. Such project-level incrementalsolution remains an important factor in costs are borne by economic agents, usuallydecisionmaking. firms, parastatals, or government agencies.

Although the financial policy question The economic incremental cost of an ODSmay have some technical aspects (the likely phaseout can be calculated, usually by ad-response to incentives, for example), it will justing financial cost data to remove transferalso be influenced by nontechnical consider- payments and the effects of distortions.ations. The Global Environment Facility Typical adjustments will be for taxes andgoverning committee and the conferences of duties, subsidies, and excess profits due tothe global environment convention parties structural imperfections.will have to consider the nontechnical con- The sector-level incremental cost arisessiderations when they deliberate on the when both production and user sectors exist,meaning of "agreed full incremental cost"' and it is necessary to consider sectorallyor "agreed incremental cost"2 as the basis balanced investment programs. Here afor international grant financing. phaseout in one sector that is not matched in

Third, we distinguish between financial the other sector could increase overallcosts and economic costs. Financial costs phaseout costs. In such a case, the invest-are relevant to enterprises and to the design ments in production and user industries areof incentives for managing a phaseout. not independent. Sector-level incremental

costs were not explicitly considered in our

1. See United Nations Framework Convention on original taxonomy, mainly because oDsClimate Change, Article 4.3, and the Convention on phaseout is largely a cumulation of projectsBiological Diversity, Article 20.2. by individual enterprises and sector costs

may not differ substantially from the sum of2. See Agenda 21, Ch 33.16; Montreal Protocol (as project costs. Sector costs in fact are betteramended) Article 10.3 (a); and Global Environment illustrated in other contexts such as powerFacility (GEF) Instrument, paragraph 2.


system investment programs where system For other global environmental problems,effects and system costs are more pervasive. these assumptions may need to be reconsid-In the power sector, it is seldom possible to ered. For example, greenhouse gas emis-consider project-for-project substitutions; sions will not be phased out, but maintainedinstead, it is necessary to consider alterna- at reduced levels indefinitely, and policytive integrated investment and operational responses will be much more important.plans. The sector approach also allows us to However, even for this problem, there mayadd policy alternatives to go beyond the still be a need for bottom-up modeling ofsimple addition of projects, which is the short-term responses that provide informa-focus of the Montreal Protocol. Such policy tion about investment options. In fact, veryalternatives could include industrial strategy few top-down models have been applied tofor ODS phaseout, and carbon taxes for this problem in developing countries.greenhouse gas (GHG) reduction. The For bottom-up calculation, we explicitlycountry-level incremental cost will include consider the level of services provided bynot only the project, or sector, costs but also ODssand ODS technologies and examine thewelfare losses and macroeconomic costs. incremental cost of alternative (non-ODS)

Originally we distinguished between technologies for providing the same typeexpenditure and loss. At the project level and level of service. If we used an equiva-this might be a useful distinction: the higher lent approach to the climate change prob-cost of substitutes and capital plant versus lem, we would hold the services of thethe loss due to accelerated replacement of power sector fixed while examining alterna-capital plant. However, at the country level tive (less GHG-intensive) forms of supplythe distinction disappears, because it is and demand management.necessary only to compare one overall coststream with another. All economic loss due The Incremental Cost Frameworkto accelerated replacement will be implicit. We calculate incremental cost using a

bottom-up approach. Whether one is calcu-

Cost Calculations lating project, sector, or country-level costs,the broad framework is the same.

Ideally, the comparison of with and with-out cases would be performed in a general Scenariosequilibrium framework to capture all the It is critical that we first identify thetrade-offs and the full adjustment costs to scenario. The scenario is the set of all as-the economy. For the ODS phaseout prob- sumptions that are regarded as fixed-thatlem, this is generally unnecessary and may is, not part of the strategies being evaluated.also be impractical. The assumptions arethat it Is unnecessary because (a) the overall Whlastteydernsteateatvadjustment cost is small relative to the course of action and is controlled by thegovernment or other economic agents, awhole economy; (b) the phaseout takes scenario, by definition, is not controlled butplace over a relatively short time (about ten describes the other relevant features of theyears, possibly less); (c) the costs are to be issue. The same scenario would thereforecompensated at a project level as they occur;

and d) tat cuntres gneraly wuld ot ~underlie both the baseline and the alternativeexpected to change overall economic policy cases described below. It is important not toor industrial strategy in order to minimize confuse the scenario with the strategies thattenefor industrialstraternationalr tompens y are being evaluated, because to do so canthe need for intemnational compensatory lead to error-such as comparing one strat-payments or to meet their commitments.

* ~~egy to another under different scenarios.

6 OZONE LAYER PROTECTiON: COUNTRY INCREMENTAL CoSTS

Different scenarios, representing different ing policy, and the "optimal" baseline,possible views of the future, can be used which includes assumptions about whateverthough to test the sensitivity of the incre- pricing policy reform and industrial restruc-mental cost calculation to uncertain parame- turing would be suggested by the county'sters, but in each case they should be applied own best interests. The impact of this choiceto both the baseline and the altemative could be substantial, because many econom-strategy. ically attractive but unfinanced opportunities

could also benefit the global environmentDomestic benefit (such as energy conservation, in relation to

It is necessary to identify the domestic the global benefit of reduced GHG emis-

benefit of the activity that is to be substi- sions).tuted or modified. An example of domestic The choice of baseline is restricted by the

tuted or modified. An example of domestic financing policy in the case of the Montrealbenefit iS the effect on the domesticohealt Protocol, but is less specified in the case ofand economy caused by the production of a the FCCC.certain number of refrigerators of given nte cael

capacity. ~~~~~~~In the case of the Montreal Protocolcapacity. financial policy, the baseline benefits are

Baseline further limited by being described as "basicdomestic needs" in the amount that existed

It is then necessary to construct a baseline at the time the Protocol was signed (1987).case. Some baseline cases may be trivial, in These lirnitations can be interpreted to meanthe sense that no activity would normally that the production of ODs-dependent goodstake place in the absence of global environ- for export is not included in the baseline andmental considerations. Others may be more that growth in the demand for oDs-depend-complex, and involve a program of invest- ent goods over time is not included beyondment and operations over a long period of the point where additional capacity is re-time. Since the baseline will be counterfac- quired.tual once an alternative has been imple- In the case of GHG reduction, many devel-mented, it is even possible to consider vari- oped countries are working towards targetsous baselines corresponding to different of the same form: emissions by a certainscenarios, for example. It will be useful to target date (such as the year 2000) will bemake the underlying assumptions explicit, limited to an amount equal to historicalsince some baselines may be more plausible emissions in a base year (such as the yearor have more policy relevance than others. 1990), with future emissions reduced by aFor instance, if the financial policy requires fixed percentage of these (possibly 20 perit, the baseline may need to incorporate any cent by the year 2020). However, develop-of the following: (a) the most economic way ing countries have no specific commitmentsof achieving the identified domestic bene- under the Fccc. It is also likely that GHGfits, (b) consistency with national priorities, emission reductions will be measuredand (c) reasonable environmental protection against normal trends rather than base yearstandards. Alternatively, it may simply be emissions in order to provide for continuedthe most likely outcome, including any industrialization to increase living stan-inefficiencies, distortions, or non-economic dards; that is, the baseline would includeconsiderations that the country is likely to "normal" growth in emissions over time.have permitted.

In general, one of two baselines is usuallyimplied: the "business-as-usual" baseline,which represents the continuation of exist-


Technical options tal costs may be sufficient for reimburse-ment if this is also done as the costs areIt IS necessary to identify technical op-

tions for producing domestic benefits of the incurred rather than in advance.same type while at the same time providing Global benefitglobal benefits.

It may be useful to calculate the incre-Alternative mental global benefit, as is shown symboli-

An alternative case needs to be produced cally above, even though it is certainly not awhich yields domestic benefits in the same part of the calculation of incremental costamount as the baseline case which it is and it is usually not monetizable. In ODS

phaseout, this is the amount of ODS phasedreplacing. This means, for example, that in out-also, thi e prof nt of rDuctioneithe ODS phaseout case, the same number of isthe GHG reduction; orefrigerators or, in the global warming case, isdiversit cervation it the conse rthe same amount of electricity will need to tionglet. tissefulto calculte gobal

be poducd bythe ltemtivestraegy s is tion goal met. It IS useful to calculate globalproduced by the baslnae strategy. benefits in order to make cost effectiveness

comparisons. If a higher incremental globalbenefit is derived from the use of givenIncremental costfunds in one activity or country than in

Incremental costs (whether at project, another activity or country, it benefits thesector, or country level) are calculated as the global environment more to devote thedifference in cost between an alternative and resources in the most cost-effective manner.a baseline case within the chosen scenario.A simple framework is provided in Table Cases1-1. There are many possible cases that can be

These costs will be distributed over time. evaluated depending on how many scenariosIn some years the incremental cost will be (s), reflecting external uncertainties, need topositive, but in others it may be negative. be considered; how many baselines (B) areFor example, in phasing out ODS in air-conditioning, there may initially be higher oiteres to olicy-akes; anhow manyincremental capital costs due to the higher alteratives (A) o aties, aundercosts of substitute chemicals. On the other consider atio.Temxu numberhand, in recycling ODS there would be incre-mental capital costs that are partially offset = x B x Aby future savings due to the lower volumesof ODS required. Likewise, in GHG reduc- The actual number of cases that needs to betion, capital costs of the alternative fuel may considered may be smaller if certain combi-lead to initial incremental capital costs of nations are regarded as unlikely or unneces-fossil fuel replacement. These costs would sary.be partially offset by future negative incre- The number of alternatives itself dependsmental costs due to the avoided baseline on the number of phaseout schedulescosts of the fossil fuel replaced. Often, the (D)-each of which must at least meet anyincremental cost is expressed as a single broad commitment under a convention-andnumber: the net present value calculated on the policies (P) under consideration:the basis of an appropriate discount rate.This is not always a necessary step, how- A = D x Pever, since the time profile of the incremen-


There is no prescription for the number that benefit that can occur in the incrementalmight be considered-this will be a matter cost framework.of judgment in each case. In the section * First, there is the baseline domesticbelow entitled Lessons for Other Global benefit-the benefit served by the activityEnvironmental Focal Areas, we consider the damaging to the global environment. Thisnumber of cases usually adopted for the benefit need not be monetized, but shouldwork on country incremental cost of ODS be clearly identified and quantified sophaseout. that one can demonstrate that the pro-

posed alternative produces at least thesame benefit as the baseline it replaces. If

It may sometimes be useful to identify this were not the case, the country couldincremental domestic benefits, if any, that hardly regard the proposed activity as anaccompany the actions taken principally to alternative. Numbers of refrigeratorsbenefit the global environment. Although produced is an example of a quantifiedcare would be taken to match the baseline but unvalued baseline domestic benefitand alternative cases in terms of the baseline that is maintained in the alternative case.domestic benefits they produce (the situation * Second, there are avoided baseline costs.shown in Table 1-1), there occasionally may These are not benefits in the strict sensebe such additional domestic benefits that are at all, but are sometimes so regarded. Innot separable in this way. fact, they are an integral part of the calcu-

In ODS phaseout, there are few incremen- lation of incremental cost. The savings intal domestic benefits; one of these few is the ODS resulting from an investment innonseparable benefit of having testing facili- recycling is an example of an avoidedties for non-ODS equipment that will also baseline cost, which must be valued andprove useful in general product testing. In used as an input to the calculation of theGHG reduction, the reduction of domestic incremental net cost of recycling.pollution loads (below the requirements andnational priorities already built into the * Third, there are incremental domesticbaseline) is an incremental domestic benefit benefits. These are "side benefits," bene-that can be expected from reduced use of fits that are not part of the baseline andfossil fuel. In conserving biodiversity, the therefore not a national priority. Suchavailability of increased biological resources benefits do not usually exceed the incre-for sustainable industries is a possible incremental cost of the pioposed action; onmental benefit. As is the case for incremen- their own they do not justify the actiontal global domestic benefits, incremental economically. Nevertheless, if thesedomestic benefits are not part of the calcula- incremental domestic benefits are tangi-tion of incremental cost. However, the ble but inseparable from the global bene-financial policy for sharing the incremental fits, the financing policy may requirecost may specify that this incremental do- some amount of cost sharing between themestic benefit needs to be taken into ac- international financing mechanism andcount in some way. the country. This is an issue of financial

policy.

Benefits * Fourth, there are incremental global bene-

Different types of benefits have been the fits. These need not be monetized, and dosource of some persistent misconceptions. not form part of the calculation of incre-Therefore we restate briefly the types of mental cost.


Country Incremental Cost of Phasing Baseline

OUt ODSs For this work, only the "business-as-

Scenarios usual" baseline is used because of the re-

Typically, a scenario will include the fol- strictions imposed by the Montreal Protocol.The Montreal Protocol, for example, effec-

lowing assumptions: tively stipulates that the baseline (and the* the growth in demand for the services alternative) case include existing economic

provided by ODS or its substi- policy and industrial strategy, although this

tutes-which forms the basis for estimat- presumably does not exclude policy instru-ing the baseline domestic benefits. Often ments specifically targeted to the phaseout

high and low growth scenarios are used; (what we have termed National Ozone

* the cost and availability of technical Policy).options; and In the case of GHGs, the importance of

.. th ot n vilbltyo DadOS policy responses is likely to be much

* thechnlogest and availability of OSandODS greater. For one thing, efforts to control thetechnologies. release of GHGs will span many decades and

Usually, the scenarios below (two or may even be indefinite. Long time horizonsthree) have been defined only on the basis of make it more likely that baselines will even-

the principal country-specific uncertainty: tually incorporate economic policy reforms

the growth in demand for ODS services, and industrial restructuring. For another, the

often linked to the rate of economic growth. costs of GHG reduction strategies are likely

to be much higher than those of ODS phase-

- High-growth scenario out and inefficiencies are less easily af-

= Reference growth scenario forded. We return briefly to the question ofpolicy responses in the section Incremental

* Low-growth scenario Costs of Economywide Policy Responses.

The other factors are international ratherthan country-specific and could also be used Technical optionsto define scenarios and standardize assump- In the case of ODS phaseout, there is a

tions with other groups working on similar wide variety of non-ODS options for the

questions. Usually, technical availability is existing applications (in refrigeration and air

agreed on the basis of expert technical opin- conditioning, foamn blowing, aerosol propul-ion (such as the Ozone Operations Re- sion, solvent cleaning, and fire extinguish-

sources Group [OORG]) and alternative cost ing). 3

assumptions are used in sensitivity testing

only around the reference scenario. Alterative strategies

Domestic benefits Up to six alternative strategies could beconsidered, depending on both the schedule

Benefits are identified in each of the user

sectors. Although production is an integral

part of the chain of technical processes forproviding these benefits, care should be 3. The options are constantly changing as a result oftaken not to "double count" by including technical developments. Technical assessments areproduction of ODS as an benefit additional to made by bodies such as the Ozone Operationsp roduction of ODS as an benefit additional to Resources Group, which advises the World Bank in

that of consumption. its role as an implementing agency of the Multilateral

Fund.


chosen and the policies adopted. The sched- can calculate an incremental cost in re-ule could be characterized as: spect of each of these policy options

* Allowable. In this schedule, actions are above (King and Munasinghe, Sectiontaken at the last possible opportunity IV).consistent with an environmental com- * Minimum incremental cost. The cost ofmitment. actions that follow the efficient policy

* Accelerated. In this schedule, actions are will be the lowest incremental cost oftaken as soon as they are technically meeting the commitments.feasible in order to minimize ODS emis- * Compliance cost. This is the cost of ac-sions by reducing them below the maxi- tions that incur the lowest incrementalmum permitted by the commitment cost subject to the constraint that eco-schedule agreed to by the country when nomic policy and industrial policy are notratifying the Montreal Protocol. altered. (This constraint is imposed by the

* Optimal. In this schedule, actions are Montreal Protocol.) However, efficienttimed to meet the commitment at lowest specific measures (National Ozone Pol-incremental cost. (To determine the "opti- icy) to assist the phaseout are assumed toma e' strategy, one actually considers a be in place. In general, because of thelarge number of strategies and, through additional constraints, the complianceconstrained optimization, selects the cost will exceed the .inimum incremen-minimum cost solution.) tal cost.

The policy options could be characterized * Adjustment cost. The actions that areas: actually taken, whether efficient or not.

Therefore the adjustment cost will gener-* Efficient. An efficient (comprehensive) ally be greater than the compliance cost.

combination of policies. This policyoption was not considered because of the Casesrestrictions of the Montreal Protocolwhich explicitly ruled it out. * Scenarios (s). Generally three scenarios

are considered: high growth, reference,* National Ozone Policy. This policy op-anlogrwhOteucraiisae

tion comprises an optimal set of specific treand low growth. Other uncertaunties arepolicy instruments dedicated to the treated as addctelnar r senaos around the

phaseout of ODS but set in a general pol-

icy context of existing industrial strategy * Baselines (B). Only the "business-as-and economic policy, either or both of usual" baseline is used.which may be very suboptimal. * Alternatives (A). Three schedules (D) are

* None. This option means no new policy considered: accelerated, allowable, andis adopted. In general, this would be an optimal. Only one policy (P) is assumed:inefficient "response" that would increase National Ozone Policy, which does notthe incremental cost of the phaseout. include changes in economic policy or

industrial strategy.For ODS phaseout calculations, three

schedules, combined with only the National * Cases (c). On this basis, the number ofOzone Policy strategy, were used, making a principal cases considered is nine:totalofthreealtematives. C=SxBXA=SxBx(DxP)=3xIx3=9

* Incremental cost. For each of the chosen Since there is only one baseline considered,scenarios, baselines, and schedules, one we can simplify the naming of such cases


using a two-dimensional matrix, as shown in nornic incremental cost subject to con-Table 1-2. straints defined by the obligations under the

Montreal Protocol and the fixed policies. InTerminology essence, the problem is the same as that of

At this point we need to alert readers to developing a least-cost investment programfor the power sector. ODS phaseout costs,

the varuous uses fisomecoute ters we like power supply costs, need to be evalu-haveud above;tfirst, ba su may ated in a time dimension. There are, how-lead to misidentification of supposedly evrsoepcilfausofteDSac-comparable cases when different studies are ever, some special featurcs of the ODS calcu-compared, and second, because the loose lation.'application of some terms in the incremental Disaggregationcost framework can lead to confusion.

First, we have reserved the term scenario The demand for the services of existingto apply only to the context of the prob- ODSs must be disaggregated by ODS. Therelem, not the outcome or the strategy. The are seven groups of controlled substancesscenario embodies the relevant exoge- defined by the Montreal Protocol, each withnous parameters that are not subject to its own chemical properties, type of use, andpolicy control and which remain the same sector of application. The solution is thewhatever strategy is adopted. There may sum of up to seven separate reduction strate-be various scenarios, however, that re- gies, since the Montreal Protocol does notflect the uncertainty about the external permit trade-offs between the schedules ofcontext. the different groups.

- Second, we have used baseline in the Data adequacysense required by incremental cost analy-sis to denote a strategy which does not There is a special difficulty with data.take global environmental considerations ODS phaseout is unlike the case of powerinto account. and water development, where there are

reasonably accurate demand forecasts ands Third, we have used reference to qualify good technoeconomic data for costing alter-scenario. It could also be used to qualify ntv upyotosa ifrn clsobaseline, since several baselines are pos- operation. ln ODS phaseout:sible.

- demand patterns and price elasticitiesSpecial Features of the Incremental have not been extensively researched;Cost Analysis of ODS Phaseout

There will be a number of special features 4. We originally regarded the choice about whetherinvolved in calculating the incremental cost, to include the imposed costs of the Montreal Protocolwhich, as indicated above, will be a compli- as a special feature. However, it seems worthance cost. preserving the distinction between reimbursement of

The compliance cost is the minimum nmitigation costs and compensation for imposed costs.The coplianc cost s the inimumThe scenario would thus be that the Montrealadjustment cost given the existing economic Protocol has been adopted by the rest of the world, atpolicy and industrial strategy. In principle it least to the extent that this will affect the internationalcan be calculated from a least-cost plan of prices and availability of ODS, ODS technologies, andadjustment comprising an optimal National the ODS substitutes. At the time, we also noted thatOzone Policy and an optimal investment least-cost planning required that benefits be kept the

same, although we are now less inclined to regardplan. The principle is to minimize the eco- this as distinctive. Keeping benefits the same is also

essential for incremental cost analysis.


* rapid technical change will have a large phaseout studies but also, to the extent thisbut unknown effect on phaseout costs is possible, for studies on other global envi-within the planning period; ronmental issues.

* costs of substitutes from sources outside The Environment Department thereforethe country and from domestic producers commissioned four consulting firms toaffiliated with multinational corporations describe, without any additional research,are not easily obtained; and their applications of the methodology. The

country observations were for the purpose of* the reductions in cost due to learning and testing the methodology, and not for evalu-

scaling up of operations are not well ating specific country situations or recom-known. mendations. The consulting firms and the

countries they selected appear in Table 1-3.Distributed costs

Unlike in utility planning, the costs of Steps in Calculating the IncrementalODS phaseout will be highly distributed Costamong a variety of government agencies, Scenariosproducers, user industries, and consumers.

Essentially, the scenarios used by the case

Matched pair calculations studies featured growth assumptions for theservices of ODS based on "unconstrained"

Since the issue is incremental costs, the ODS consumption. In some cases, the sce-calculations have to be done in matched nario explicitly assumed the ozone issue hadpairs of baseline and alternative cases, and never evolved into the Montreal Protocol;not singly. that is, there were no restrictions in the

country or anywhere else. The main impactCase Studies of this assumption was that ODSs would be

available at present real prices, whereas, inBackground fact, prices might eventually rise5 as produc-

Since the theoretical work had been done, tion falls below the levels for economies ofa number of Country Programs had been scale, and international supplies mightprepared for the Multilateral Fund of the become unavailable. Strictly speaking, aMontreal Protocol. These programs assist scenario should set out the expected contextthe preparation of individual projects which for both the baseline and the alternativewould be eligible for financing by providing cases so that the incremental cost of deci-an overall program of which they form part, sions to implement the Montreal Protocol inand outline the recommended National the given country can be measured. Includ-Ozone Policy. ing the "no Protocol at all" situation in a

Our concern here is not with the country- scenario, it becomes unusable for this purespecific recommendations as such, but with calculation since the difference in cost willthe feasibility of the methodology (particu- also include what we termed the "imposedlarly in the light of the special features cost." However, given the uncertainty of theoutlined). The methodology is useful as a forecast, this may not have much of antest of proposed National Ozone Policy, as impact.a way of evaluating investment plans forODS phaseout, and as an indication of re-source needs for ODS phaseout. We would 5. Prices might temporarily drop if demand islike to draw lessons not only for future ODS restricted before capacity is closed.


Unconstrained demand is generally pro- application, the most plausible alternativejected on the basis of a base year consump- technologies are matched.tion and a growth factor derived from popu-lation and economic or sector growth. A Baselinetime horizon that is sufficiently long to Implicitly, the case studies used theallow the phaseout and all consequential "business-as-usual" baseline.costs to be adequately represented (20 or 25years) was used. High and low forecasts Technical optionswere offered, sometimes with a reference orcentral estimate as well. In each country, it was necessary to make

The main practical problem, as expected, judgments about the technical options,was the unavailability of data even for the whether these were straight chemical substi-base year. This situation arises because it is tutes, such as liquefied petroleum gas as anecessary to have the data disaggregated by propellant and HFc-134a as a refrigerant;a phaseout group, and because it is desirable process changes, such as low-chlorofluoro-to have it disaggregated by application in carbon (CFC) foam blowing techniques andorder to make better projections. As noted the use of recycling and recovery tech-by Bendtsen (in this volume) in the case of niques; or product substitutes, such as me-Tunisia, there are numerous minor applica- chanical pumps instead of aerosol sprays fortions and many small users, particularly in cosmetic applicators and cardboard insteadthe solvents sector but also in the refrigera- of foams for packaging. However, there istion sector with its multifarious servicing considerable uncertainty about which of therequirements. Import statistics and company technical options included in the preparationdata are not easily obtained, and surveys, of the alternative strategy will stand the testwhich are time consuming, are often needed. of time. The search for feasible alternativesAnother problem is that the base year itself is extensive, and recent successes havemay have to be adjusted for unusual circum- reduced the estimated global cost of thestances, such as those that prevailed in the phaseout. Furthermore, some options arelate 1980s and early 1990s in the former acceptable in one country but not in anotherCzechoslovakia and other central European (for example, methylene chloride for flexi-and former Soviet republics when the earlier ble foam blowing), and transitional sub-centralized planning and the Council for stances themselves will have to be phasedMutual Economic Assistance trading sys- out one day. There is some disagreement ontems were disintegrating. the starting dates for new technologies

Because the forecasts are not precise or (although this could reflect country avail-certain, they cannot be used for direct reim- ability rather than uncertainty about interna-bursement purposes. The actual develop- tional availability), market penetration rates,ment of demand would need to be reviewed and reduction potentials.over time for this to be assessed more accu-rately. Alternative

Domestic benefits The alternative strategy is constituted bya sequence of investments in the technical

The proxy for the benefits of ODS ser- options that are required or encouraged byvices, which are held the same for both the the National Ozone Policy. As required, thebaseline and the alternative, is the uncon- economic policy and industrial structure arestrained ODS consumption. For each ODS assumed to stay the same. Thus the savings

that might accrue if small-scale, inefficient


production were to shut down in the transi- stream. This would not be possible to do intion are not included in the final calculation. other focal areas if the alternatives are very

Accelerated, allowable, and optimal different or where there are system or mac-schedules were tested. In the case of Tuni- roeconomic costs involved.sia, the accelerated schedule was in any case However, in the ODS transition, there willonly slightly more expensive than the opti- be some economic losses in the form ofmal (least-cost) one, despite the fact that it is premature retirement of capital equipmentwell ahead of the formal Montreal Protocol and consumer appliances and also somelimits.6 benefits from the transition in the form of

more modern equipment that is acquiredIncremental cost earlier than it would otherwise have been. It

In the suggested framework, the total is this aspect of the ODS phaseout problemcosts of actions are recognized where and that the unit abatement cost approach doeswhen they occur in both the baseline and the not really cover. In the case studies, there-alternative cases. The incremental cost is fore, this issue was handled separately.calculated only once, as the "compliance Bendtsen (in this volume) argues that thecost," when the total cost of the baseline is cost of forced early retirement is extremelysubtracted from the total cost of the altelna- sensitive to the assumptions about the futuretive. However, in none of the case studies availability of drop-in substitutes as well asdid the analyst use this approach. Rather the availability of ODS from recycling andthan two cases with data on total cost being reclamation. While the forced early retire-e>ntered, the analysts used a single "increment costs are be estimated for domesticmental" case with estimates of unit abate- refrigerators, they are ignored for otherment costs (in otherswords, unit abate - equipment because the costs are probablycosts) multiplied by the amount of ODS small (due to retrofitting) and because therereplaced. is insufficient data available on the stock of

This approach, using unit abatement equipment and its age and characteristics.costs, mapproach, reasingounably a me nt fThe estimate of the early retirement costs for

costs, may work reasonably well for ODS domestic refrigerators is based on the stockphaseout, which is calculated from the and age rofile of refrigerators the annualbottom up and involves costs that are small ndu the grae of rechars, the aver-relative to the economy as a whole. Much of production, the rate of recharging, the aver-the cost data is already in an "incremental" age useful life of refrigerators, and the eco-form: capital conversion costs, additional nomic cost of refrigerators (taken to be priceroyalties and costs of patents and designs, less tax).the full costs of required technical assistanceand institutional strengthening, and incre- Casesmental unit costs on chemical substitutes. The country studies either refer to explicitAs Widge, Radka, and Dillon (in this vol- cases or implicitly use three alternativesume) point out, care is needed to prevent with three scenarios, the latter perhaps in the"double discounting" the unit abatement form of sensitivity tests around a referencecosts when they are put back into a cost scenario based on a central estimate of

unconstrained demand.

6. The Country Program recommended that the Global benefitaccelerated schedule be adopted because the globalbenefits are substantially larger for little increase in The incremental global benefit will be thecost, and because possible CFC price increases will physical reduction in the ODS emissionmake this identical to the optimal schedule. achieved by the alternative over the baseline


for each scenario. This could be expressed * NIK = not-in-kind substitutes for ODSas a time profile of the reductions of the * Subs = ODSs replaced by in-kind alterna-different ODSs. There is an unresolved issue tivesabout the means of aggregating differentODSs through a weighting factor such as the * New = new applications requiring ODSozone depletion potential and of expressing substitutesthe value in present value terms (for exam- To meet the restrictions of the Montrealpie, through a discount rate). The issue is Protocol, New may need to be zero.unresolved because there is no agreed dam-age function for the emission of ODSs. This Incremental global benefit =

is not a practical issue because the Montreal y YODP(ODS,) x[EB,(0ODSj,t) - EA,(ODSj,t)]Protocol specifies benefits in terms of the i tschedules; nor is it a serious economic issue where: i = index for each ODS group, andin the present context, since global benefit is t = index of time (e.g., years)not a part of the incremental cost calculationand there is an implied commitment to Modeling Issuesphase out all ODSs. So generally, if approxi-mate comparisons are needed, it would be Optimiizationsufficient to calculate the accumulated One of the models we used in this studyozone-depleting potential- (ODP-) weighted (Bendtsen, in this volume) was an optimiza-but undiscounted ODS emission reductions tion model, whereas the others used iterativerelative to the baseline. One could, of simulation. Optimization is an extremelycourse, determine the sensitivity of the powerful technique for testing ODS phaseoutresults to different discount rates. strategies (schedules and policies), including

Once the schedule of actions has been those for the phaseout of transitional sub-determined, this can be done by first calcu- stances, such as hydrochlorofluorocarbonslating for each ODS the annual emissions in (HCFCs). However, optimization may now bethe alternative case, subtracting the emis- unnecessarily sophisticated for the ODSsions of the baseline, and then totaling the phaseout calculations, since the problem issums of the differences over all the years highly constrained.and ODSs. The baseline emissions are the First, for those central and eastern Euro-estimates of the scenario, or, in other words, pean countries and former Soviet republicsthe unconstrained demand. that will not be accorded the status of devel-

oping countries, the actual difference be-EHl,,, = Unconstrained Demand tween the alternative phaseout speeds has all

The emissions of the alternative case are but disappeared due to the Copenhagenfoundhby remsions the al nternatined cse- aAmendments to the Montreal Protocol of

found by reducing terunconstre de- November 1992. The following amend-mand by an amount determinedxby the te- ments significanty tightened the require-

diced optissions. of Catanhe atex veseste i- ments: the required phaseout date for CFCs,

the following formula: methyl chloroform (McF), and carbon tetra-chloride (CTC) is January 1, 1996, and that

EA = EB - Conserve - NIK - Subs + New for halons January 1, 1994. There are thusfew alternatives but the accelerated phaseout

where: schedule. For those developing countriesthat depend on imports of ODs, the phaseout

* Conserve= materialsavingsfromrecover of ODS in the developing countries willand recycling.


effectively raise the price and reduce the eventually be passed on to consumers inavailability of ODS along the developed the form of higher prices.

countryi schedules. * It brings out the Issue of importationIn fact, three of the studies showed that versus production.

the accelerated schedule was very close tobeing the optimal. * It integrates supply with use. The produc-

tion, industry use, and consumer use ofModel Sophistication ODS and ODS appliances form an inte-

grated system. Where there are onlyThe models used in the case studies could independent uses, such uses can usually

be run on personal computers, and both be evaluated separately in terms of theirCatanach (in this volume) and Miason (in individual cost-effectiveness. But wherethis volume) mentioned that their models they are all linked through an integratedwere spreadsheet based. In view of the supply chain, the analysis is more compli-requirements of the model, and the need for cated. If the phaseout is not balanced, thetransparency, this seems an excellent choice. rapid phaseout in one sector will imposeModern spreadsheet programs offer good costs on the others. For example, phase-databases and scenario managers that facili- out of production, assuming that importstate computation with conceptual under- cannot take up the slack, will imposestanding. costs on user industries which face a

sudden rather than a gradual withdrawalSupply-Side Modeling of ODS.

Mason describes experience with the * It allows the systematic exploration of themodeling of a large integrated ODS system dominance of the costs of the early retire-(India) which has production, user indus- ment of appliances when CFCs are notries, and consumer applications. Simplified longer available for recharging (a factapproaches were taken for countries with noted by both Catanach and Mason). Thisfewer elements, such as Turkey, which has dominance suggests that an accelerateda wide range of industrial activity that uses schedule is preferred, especially in grow-or supports the use of ODS but does not ing economies, because it limits the accu-manufacture it, and Zimbabwe and Jordan, mulation of capital stock and consumerwhich also import the components used in durables that would eventually requireequipment that contains ODSs. replacement. However, an accelerated

In general, though, the supply side was schedule may impose high initial costs ofmodeled less effectively than the user indus- replacement, unless these can be con-try sectors. This is understandable for most tained in some way. This cost contain-developing countries, which are importers ment usually requires one or more of theand users, but is more serious when there following supply-related options:are major supply side options. The mainreasons for having a good supply side model * that CFCs be recovered, recycled, andare the following: reclaimed in order to provide a continu-

ous supply of cFc for appliance servicingIt helps to identify costs where they first but without the necessity for importing oroccur. In an economic model, one is not producing them;interested in the incidence of the costs; infact one assumes in this case that costs * that drop-in substitutes are readily avail-originally borne by the producers will able; or


that retrofitting can be done easily and Bendtsen regards a doubling of CFC prices ascheaply. sufficient for this; Mason believes that the

A model that adequately represents these accelerated schedule already reduces thesupply-related options is needed if genuine phaseout costs in India significantly and thatleast-cost planning is to be accomplished in in other reasonably open economies, likethe more complex economies such as Chile, Jordan and Zimbabwe, the phaseout will beIndia, and China. The costs of these options market-driven by its links to the acceleratedshould be considered in the overall cost phaseout in supplier countries. Widge notesminimization. that changes in the prices in ODS and substi-

tutes attributable to the global phaseout will* It allows further exploration of assump- have substantial impacts on Thai industry

tions about, and actions to affect, the and consumers; however, given the mixedservicing of equipment. On the one hand, baseline used (which assumes a "no-proto-Widge suggests that the simulation of a col" situation) the results in the Thai model"servicing tail"-that is, the use of ODS are likely to be inconclusive. The rising costbeyond the date of the phaseout in new of ODS will of course add to the costs ofequipment-is useful and was done for industry, but these are strictly baseline costs;the Philippines and for the island of Tai- the incremental costs of the substitutes willwan (studies not included in the case in fact decline in this situation.studies here). On the other, Bendtsen Although it seems that the recommendednotes that for Tunisia the result (choice of strategies are robust, most of the case studyphaseout speed) is not sensitive to the authors call for more research on the pro-assumed rate of servicing of domestic jected costs of ODS and ODS substitutesrefrigerators. Perhaps the issue to explore because this will have a major impact onfurther is not so much the choice of over- incremental cost. Bendtsen notes that theall phaseout schedule (which in most costs of ODS substitute technologies tend tocases seems robust), but the choice and become generic if world prices are used forcost-effectiveness of subsidiary strategies the open economies: projected worldwideto further reduce the residual uses of ODS costs would therefore help project prepara-once the Montreal Protocol targets have tion, being both cost-effective and standard-been met. ized. Catanach comes to a similar conclu-

sion, and calls for identification of the sup-Further Research Effort ply costs (freight costs can then be estimated

Further research needs can be divided and added). Bendtsen also notes that operat-into studies of ODS phaseout and those ing costs tend to dominate the costs of theinvolving the mitigation of other global phaseout (except in the "allowable" sce-environmental problems, which are less well nario, where the costs of forced early retire-defined. In relation to the ODS phaseout, we ment dominate, as expected) implying per-can consider the following findings and haps that it is the future costs of substitutesensitivity analyses. chemicals in particular that are the important

parameters.International price data While generic information would of

course be useful, would help standardizeA major increase in ODS prices (as would country studies for comparative purposes,

seem likely as plants close) makes the accel- and would be most cost-effectively gener-erated schedule the least-cost one, thereby ated if prepared centrally (for example, formaking it less necessary to study this aspect the Multilateral Fund), we doubt that thisfurther with optimization techniques.

18 OzONE LAYER PROTECTION: COUNTRY INCREMENTAL COSTS

alone would be as helpful as believed. As most cost-effective or close enough to it thatnoted above, most of the case studies used it is preferred as a risk management strat-unit abatement cost approach in their analy- egy.) That is, although the country incre-ses, and this may have inappropriately sug- mental cost itself is difficult to calculategested that incremental costs can be mea- with any confidence, the framework andsured or projected on a purely generic basis. process yields important general results.In our experience (World Bank and Mon-treal Protocal 1992), the unit abatement Phaseout policycosts are far more sensitive to country-spe- Typical measures constituting the recoi-cific factors such as the utilization of capital mended National Ozone Policy in the stud-

equipment than on generic factors. Mason in ies were: taxes, selective bans, standards,fact makes an observation that seems to I

monitoring, joint government-industrysupport our conclusion in relation to recy-suort u a . committees, and information dissemination.cling:ststherecycis gin a majorgdi e between shopAlthough Mason states that country incre-the costs of recycling in a large repair shop, mnacotosietosmygudthwith centralized repairs, and those of smallservice shops that cater to on-site repairs. tiring of such measures, it was generally

conceded that the measures themselves werenot modeled explicitly but recommended ongeneral economic efficiency grounds or

An increase in country-specific growth because of their institutional feasibility.certainly adds to the incremental cost, but (Nevertheless, we expect that modeling ofdoes not seem to alter the conclusion that policy instruments will become much morethe most appropriate phaseout schedule is important for other global environmentalthe accelerated one. Catanach notes this in issues, such as the phaseout of GHGs, be-the case of Egypt; Bendtsen notes only that cause of the longer time horizon over whichan increase in the growth rate raises the the reduction takes place, the number andincremental cost. This suggests that further diversity of sectors in which action will bework on refining country growth factors will taken, and the pervasiveness of the mea-not help in strategy formulation, and since sures.)country incremental cost is not the basis ofreimbursement, it would not help in estimat- Investment sequencinging resource transfers either. To the extent The models all set out the required se-that the Montreal Protocol limits reimburse-Thmoesaletutherqids-that tothe Montrealenrotol limits inrelatimbuquence of investments in a transparent way;ment to the incremental costs in relation to this was certainly a strength and by itselfcapacity extant in 1987, growth projections would have assisted the planning process.are not very relevant for anticipating re- However, the analysts did not appear tosource transfers. However, they would alert have used their models to perform an opti-the country to the extent to which mal sequencing of investments. It is likely

unfundable incremental costs are growing. that they timed the investments on the basis

Usefulness of the Approach of the availability of the options and thematerial balance required by the schedule

Phaseout schedules that formed part of the proposed strategy.

The methodology was able to demon- Bendtsen believes that the cowIconsultstrate the relative costs of the alternative model is not well suited to sequencing.speeds of phasing out ODS. (In most coun- Unlike the power sector, where centraltries, the accelerated schedule is either the Utility planning IS the norm, the various


sectors where ODS will be phased out are Incremental Cost Calculationscharacterized by the multiplicity of small,private undertakings. Competitive market Scenariosforces often drive the phaseout, and an "opti- In climate change, the scenario will in-mal" sequence of investments may not be clude such things as the demand for energypossible because it might disturb the current services, and the likely international pricesmarket shares of different companies. In for principal fuels and important technolo-many cases the data available are not ade- gies. Scenarios underlying the climatequate for determining such optimal se- change problem are probably easier to treatquences. than those for ozone depletion because

energy, which is a significant factor drivingLessons for Other Global climate change, has been the subject of such

Environmental Focal Areas analysis for reasons of commodity andresource forecasts for many years. The

Other Global Environmental Focal specific factors that should be included inAreas the scenarios underlying the other focal

areas are not well known, partly because ofIn conclsion,it ca besatedhatgobal the diversity of actions that could be in-

environmental problems are caused by volved in mitigation.actions in individual countries. The mitiga-tion of those problems will lead to global Domestic benefitbenefits-reduced crop damage, reducedrisk of climate change, and so forth. Eco- In climate change, the domestic benefitsnomic mitigation-actions whose costs are that will need to be preserved by alternativeless than the global benefits they pro- (low GHG-emitting technologies) will beduce-will nevertheless impose costs on the energy-related; others will be food andhost country that are less than that country's transport services. One lesson from ourown share of the benefits from that specific studies of ozone depletion is that there willaction. Therefore, for mitigation to take be difficulties in projecting these benefitsplace, international commitment is neces- from base years on the basis of growthsary, and-in the case of developing coun- factors. This is, first, because some basetries which cannot afford the actions implicit years will be unusual (especially those forin that commitment-international financ- central and eastern Europe around 1990,ing. International conventions provide fi- which is the commonly regarded base yearnancing for the incremental costs of such for developed country emissions) and sec-actions in developing countries in the case ond, because compound country and sectorof three focal areas-ozone depletion, cli- specific growth over long time periods willmate change, and loss of biodiversity. The be difficult to make with confidence.Global Environment Facility, in addition to The projected benefits for industries andfinancing actions in those areas, also fi- agricultural concerns that are the source ofnances actions to prevent the pollution of land-based pollution of international watersinternational waters. will be similar in type to those used in ozone

There are similarities and differences depletion because they will involve numer-among these focal areas that may affect the ous competing firms and various pollutingability to calculate the country incremental processes.cost. The important aspects are listed in The projected benefits to be maintainedTable 1-4. in biodiversity protection measures will be

particularly diverse, reflecting the diverse


nature of the interventions necessary. How- ducing menus of options for biodiversityever, in some countries, land pressure is the conservation will be less easy because somain factor in biodiversity loss, and the many options will not be straight technicaldomestic benefits of land use may be substitutions but rather local choices such asprojectible. siting alternatives and land use options.

The simple imposition of a constraintBaseline (such as a GHG emission constraint) on the

optimization of investment and operationsThebl basl in.e queson wi te chamoe will lead to incremental costs-but incre-

prdbiobl vema iti in ss thease fof clim e change mental costs that are much higher than theyand biodiversity loss than for ozone deple- nedb.Temscotfetierpneto .Thi istu,fis. bcueo h og need be. The most cost-effective responsetion. This is true, first, because of the long, will require a careful choice of technical

possibly indefinite, time spans; secondoptions and policies. This in turn will re-because there will be many more sectors

involved than just a subset of the industrial quire an assessment of the future costs ofi b t i new and emerging technologies. This has

sctos wilb mhird her.LBec the . proved to be difficult in the focal area ofcosts wilo an c ofgae playing, thi ODS phaseout; it will be even more so inscope forclimate change and other areas where trea-likely that more attention will focus on the ties and domestic olicies are encoura ingspecification of the appropriate baseline. g gspc I I * 'pd technical 'nnovatiO'The most important factor, of course, is rapi I ion.whether this will be "business-as-usual" or Incremental costbe based on the assLumption (more or lesstrue in the long term) of policy reform. In Many of the incremental cost calculationsthe case of climate change, one policy varn- for the ODS phaseout appear to beable will be most important: energy pricing. overconstrained. This implies that there are

In biodiversity loss, important policy no degrees of freedom, nor opportunities tovariables will be the pricing of agricultural explore cost-effective solutions. In order toinputs and products, land tenure, and the avoid this situation in other global environ-incorporation of domestic environmental mental issues, it would be necessary toconsiderations into decisionmaking. construct treaty targets, schedules, and

eligibility requirements more broadly.Options and alternative strategies Where responses are interconnected, inte-

In climate change (at least in the energy grated models will be necessary, as in thearea), many technical options will be simi- case of countries that have ODS production,larly dependent on technical change. Exam- user industry, and consumer applications.ples of such technical options are in theareas of new and renewable energy technol-ogies, demand side management techniques,and fuel switching. In this respect, the con-struction of alternatives from a menu oftechnical possibilities will be similar.7 Pro-

7. In current work on the power sector for the alternative power development plans that satisfy theProgram for Measuring Incremental Costs for the same domestic needs (electricity demand) but at aEnvironment, the Latin American Energy lower emission of GHGs by using differentOrganization (OlADE) is working with the power technologies.utilities of Colombia and Costa Rica to construct


Bibliography Montreal Protocol on Substances that De-plete the Ozone Layer. Concluded at

Agenda 21. 1992. Agenda 21, Report of the Montreal on September 16, 1987; enteredUnited Nations Conference on Environ- into force on January 1, 1989; text inment and Development, Rio de Janeiro, International Legal Materials, vol. 26June 3-14, 1992. UN Doc. A/CONF. (1987), pp. 1550-61.151/26/Rev. 1, vol. I, pp. 9-479. United Nations (UN). 1992. Convention on

Benedick, Richard. 1991. Ozone Diplomacy. Biological Diversity. Opened for signa-Cambridge: Harvard University Press, p. ture in Rio de Janeiro on June 5, 1992;149. entered into force on December 29, 1993;

Center for International Climate and Energy UN Doc. UNEP/Bio. Div/N7-iNc. 54Research, Oslo. 1993. "A Review of (1992); text in International Legal Materi-Country Studies on Climate Change." als, vol. 31 (1992), pp. 822-41.Working Paper No. 7, Global Environ- United Nations (UN). 1992. United Nationsment Facility. Framework Convention on Climate

Global Environment Facility. 1994. "Instru- Change. Opened for signature in Rio dement for the Establishment of the Re- Janeiro on June 4, 1992; entered intostructured Global Environment Facility." force on March 21, 1994; UN Doc. A/AC.Washington, D.C.: Global Environment 237/18, Pt. II/Add. I and Corr. 1 (1992);Facility. text in International Legal Materials, vol.

King, Kenneth, and Mohan Munasinghe. 31 (1992), pp. 851-73.1991. "Incremental Costs of Phasing Out World Bank and Montreal Protocol. 1992.Ozone Depleting Substances." Environ- Sourcebook of Benchmark Unit Abate-ment Working Paper No. 47, Environ- ment Costs. Policy Department (October).ment Department, World Bank. Washington, D.C.: World Bank.


Table 1-1: Scenario X

Domestic Global benefitCost ($) benefit ($) (avoided emissions)

Alternative CAIt DB - GAIT

Baseline Co. DB - GB.

Increment (CAI - CBR) > 0 0 (because benefits (GB, - GAJ) > 0are held constant)

Table 1-2: Baseline: Business as Usual

SCENARIO-->

ALTERNATIVE High Reference Low

Accelerated

Allowable

Optimal

Table 1-3: Consulting Firms

Firm Principal author Country cases

COWIconsult, Ulla B. Bendtsen Tunisia, CzechoslovakiaDenmark (Comments on Ghana)

Touche Ross, Caroline Mason India, Turkey, JordanUnited Kingdom Zimbabwe

ICF Incorporated, Vikram Widge ThailandUnited States (Comments on Taiwan

and the Philippines)

C. Catanach, Charles B. Catanach EgyptUnited States


Table 1-4: Country Incremental Costs of Protecting the Global Environment

<-----------------------------------FOCAL AREA-------------------------->

Pollution of inter-Aspect Ozone depletion Climate change national waters Loss of biodiversity

Imposed cost of Food chains, crops, Storm frequencies, Marine food chains Species lossglobal change immune response, sea level, droughts

skin cancer and floods

International con- Montreal Protocol Framework Con- Law of the Sea, var- Convention on Bio-vention of the Vienna Con- vention on Climate ious regional seas logical Diversity

vention Change treaties

Mitigation targets in Phaseout schedules No specific ones for None Noneconvention for each group of developing coun-

ODSs tries

Separate mitigation Seven ODS groups Carbon dioxide, Yes (organotins, Numerous (variousgroups with separate tar- methane, and other petroleum) but not ecosystems and spe-

gets GHGs formally classified cies)

Treaty financing Yes; Yes; No Yes;mechanism Multilateral Fund GEF (interim) GEF (interim)

Reimbursement for Agreed Incremental Agreed Full Incre- Agreed Full Incre-mitigation Cost mental Cost mental Cost

GEF focal area Yes; Yes Yes Yescertain countriesnot covered by theMultilateral Fund

Mitigation Complete phaseout Reduced emissions Reduced pollution Protection of eco-by certain com- systemspounds

Benefits proxy ODS GHG Various chemical No. Goals specifiedloads exogenously

Method of aggrega- Ozone depleting Global warming None Nonetion of benefits potential (ODP) potential (GWP)

Incidence of costs ODS producers; user Power and energy Farms (pesticide Government (con-industries; consum- suppliers; farms; and fertilizer users); servation and pro-ers; governments industries; transport industries (toxic tected area manage-

effluent); shipping ment); industry and(toxic cargo, ship agriculture (pollu-waste); government tion); people (alter-(oil and chemical native land use)spill cleanup)

Evaluation of Incremental Cost Methodologiesfor Phasing Out Ozone-Depleting Substances:

Case Studies of India, Turkey, Jordan, and Zimbabwe

Caroline Mason

THIS CHAPTER PRESENTS THE MAIN RESULTS of case studies ofozone-depleting substance (ODS) phaseout studies in fourcountries: India, Turkey, Jordan, and Zimbabwe.'

1. The work was carried out by Touche Ross Management Consultants in response to arequest from the Environment Department of the World Bank.

25


Before presenting these findings, it is help- were reconciled. From this data, we did theful to distinguish between two principal cate- following:gories of Article 5 countries: * prepared an inventory of the ODSs being

* countries such as India that manufacture used in the country, including the annualODSs and have a wide range of industrial consumption (in tons) of each;activity to use them; * established the applications using ODSs,

* countries that import their ODSs. These including quantities for the manufacture ofcountries can be further subdivided as new products and for recharging equipmentfollows: during repair;

* countries such as Turkey that locally * determined the source of the ODSs, whethermanufacture components (such as com- imported or locally manufactured;pressors) for use in equipment, as well * presented an outline of the industry struc-as assembling finished goods (such as ture in each sector; andrefrigerators) and using ODSs in somemanufacturing processes (such as in * estimated the stock of refrigeration and airfoam blowing); conditioning appliances in the country and

their approximate age profile.* countries such as Zimbabwe and Jordan

that also import the components for use Future Demandin equipment that contain ODSs. Theremay be some industrial activity in as- Likely future demand for each ODS use wassembling finished goods from the im- projected to the year 2010 (2015 forported components, and some use of l,1,l-trichloroethane) in the absence of anyODSs in manufacturing processes. limitations imposed by the Montreal Protocol.

The forecasts were kept simple, and wereIn this chapter, the focus on the methodol- bsdpicplyo fiilpplto n

ogy developed in India, a country with a economic growth projections. For India, bothcomplex industrial structure including ODSproducers, equipment manufacturers, and end upper and lower bounds of likely growth wereusers. Simplified versions of this methodology projected, based on different scenarios.were applied in Turkey, Jordan, and Zimba- The resultant unconstrained demand fore-

bwe, whereheresnoOSprouco. casts provided the baseline estimate of thequantities of ODSs to be substituted if demandfor products is to be fulfilled, thus maintainingMethodology Used to Calculate the domestic benefit. Given the uncertainties

Country Incremental Costs associated with forecasting, especially so far

into the future, the projections were onlyThe stages in the methodology used follow, indicative of likely future trends.

Current Situation Montreal Protocol Phaseout Profiles

Base data for India were collected by an In India, the phaseout dates expected toIndian consulting firm, drawing on govern- emerge from the London meeting of the Par-ment officials, published government sources, ties were used. In the other countries, theindustry associations, and individual compa-nies. Data was also collected independently byTouche Ross. The data varied in quality,consistency, and completeness. As far aspossible, supply and demand side estimates

2. Incremental Cost Methodologies for ODS Phaseout: India, Turkey, Jordan, and Zimbabwe 27

Substitutes and Alternatives * taxes, excise or other, or import tariffs on

Technical options, in the formi of substi- ODSs or products, or subsidies or reducedtutes and alternatives that were likely to be taxes on non-ODS substitutes or products;adopted in India, were identified for each use - direct financial support for recycling orbased on expected best practice around the recovery schemes; andworld at that time. By and large, the Indian * publicity to inform users of the changescompanies studied did not have plans for they face, and training to encourage goodphaseout. The unit costs and technically feasi- housekeeping practices.ble timing for changeover were estimatedfrom international sources, taking into account Phaseout Alternativeslikely time lags for introduction into India. Inmore recent studies, the situation has changed, The current situation, future demand, Mon-and some companies do have plans for phase- treal Protocol limits, possible substitutes andout. In these situations, we have also taken alternatives, and available policy measures,into account the existing plans of the compa- were all drawn together to produce phaseoutnies involved, and the possible influences of alternatives, or possible patterns of projectedtheir suppliers and any foreign collaborators. phaseout measured in tons of ODSs. We devel-

oped a suite of computer models that allowedPolicy Measures modeling of each user sector individually, as

well as allowing a combination of models toAt the time of the India study, we antici- provide an aggregated country perspective.

pated that government action would be needed For each user sector in India, we assessed theto drive phaseout. More recently, in countries likely response to the package of governmentwhere there are no producers of ODSs, market measures. Figure 2-1 shows the estimate offorces are proving to be powerful, while the the likely responses of the domestic refrigera-government role is more limited. We built up tor manufacturers in India to a suite of newan understanding of the industrial, environ- taxes and regulations, which provide themental, legislative, and fiscal framework in stimuli first to introduce water blowing for theeach country, and of the broad economic foam insulation, subsequently to switch to thepoiicies and strategies. We considered the hydrochlorofluorocarbon HCFC-123 as theeffectiveness of regulatory and fiscal measures foaming agent, and the hydrofluorocarbonalready used and in use. From this overview, HFC- 1 34a as the refrigerant. Finally, in thewe identified and evaluated a range of policy absence of continuing chlorofluorocarbonmeasures (both market-based and command- (CFC) 12 supply, recharging of refrigeratorsand-control) available to the government to that fail in service has to stop. The responsescause industry to take action. These included of each sector are additive to generate the

* quotas on production or consumption of country's overall CFC profile. Similarly, theODSs, either voluntary or enforceable by phaseout profile for each group of substanceslegislation; was built up.

* fiancial incentives to close ODS production The timing of introduction of the measuresaffects the timing of responses. Different

capaity to establish stu capac - timings were modeled to investigate, forproduction or use of substitute substancesy example, at the two extremes, the latest possi-

ble phaseout consistent with the Montreal* assistance with technology transfer; Protocol schedules and the earliest phaseout

that is practically possible.


Country Incremental Costs In ODS nonproducer countries, the addi-tional cost of substitute chemicals over and

The conoic csts f phseou comrise above that of the ODSS they replace was mod-one-time capital costs, ongoing incrementaloperating costs, and the costs of forced early eled as an incremental cost per ton imported,replacement of equipment reliant on recharg- rather than as the additional capital and oper-

atmns costs associated with new chermicaling with ODSs. It is assumed as a baseline that gindustry continues to meet growing demand production capacity. A discount rate appropri-for ODSs and for products made with or con- ate to the country (such as the goverment'staiing ODss by investing in additional produc- real discount rate for major investment pro-

tD ~~~~~~jects) was adopted in each case,tion capacity as required. We estimated thecountry-specific economic cost of different Preferred Alternativephaseout profiles by calculating the annualincremental cash flows in current prices be- Through an iterative process, we investi-tween the time of the study and 2010 (2015 gated what might constitute an optimal strat-for I, l,1 -trichloroethane) and presenting these egy for the government to pursue. Optimal incalculations as the net present value. We then this context meant least-economic-cost to thebuilt up estimates of incremental unit costs for country, taking into account practicalities ofeach sector (ODS producers, each of the user implementation. The phaseout timing modulesectors, and consumers). The elements of the (see Figure 2-2) was varied to calculate themodel are illustrated in Figure 2-2 and are effect on the country incremental cost. Also,described in more detail in the section on recycling options were considered to reduceModeling below. the number of refrigerators that needed early

For the ODS producers and users in manu- replacement.facture, most additional costs will, in the longterm, be passed on to the consumer of the Sectoral Implicationsproduct in the form of higher prices. However, We then worked back from overall countrythe costing model identifies costs at the point incremental costs to summarize the implica-at which they first arise, rather than at the tions for each sector. It is at this stage thatpoint at which those costs are recovered from s

customrs. T avoi doube couting,costs specific projects potentially suitable for assis-customers. To avoid double counting, costs tac frmteMliaea udmyb' ~~tance from the Multilateral Fund may beincurred by producers are not also represented identifiedin the costing model as being borne by usersor consumers in the form of higher prices. ModelinWith this approach, those incremental costs gwhich fall directly in the consumer sector and The computer modeling used Lotus 1-2-3which are additional to the incremental costs on personal computers, with inter-linkedincurred in the producer and user sectors are spreadsheets on a modular basis as illustratedthose arising either from changes in running in Figure 2-2. For maximum flexibility, wecosts of appliances or from the need to bring used simple simulation techniques, rather thanforward replacement of appliances which optimization techniques, to investigate thecannot be recharged because the recharge effects on national incremental costs of differ-substance is no longer available. Early re- ent phaseout timings.placement has been modeled as a cash outlay We first entered annual growth percentagesof the full cost of a new appliance in the year for each user sector, then developed altera-of replacement and a cash saving of an equal tive phaseout timings (consistent with theamount in the year of natural replacement. policy measures selected) and input these.

These policy measures were used to reduce


demand in the user sectors. In India, we then Country-Specific Assumptionsphased down production in the ODS producer and Resultssectors in line with these policy measures. Innonproducer countries, we assumed that ODS Indiaimports declined in line with declining de-

* ~~~~~~~~At the time of this study (January-Maymand. All prices were assumed to stay con-stant in real terms. 1990), we assumed likely substitutes and

stantInvrealtmerms. innewproducioncapacalternatives for ODSs in India as set out inInvestments in new production capacity Table 2-1. The table also shows assumptions

(whether in the producer or user sector) werebased on typical plant sizes in each country for about the earliest dates that these substitutesbased~ ~ ~ ~ ~ ~~~~~~~n alenaie woulda reaisicll bee avail-cunryfothat type of facility, even if this did not seem and alternatives would realistically be avail-to represent the optimum economic size of able in India.increments. The unit cost assumptions used are given in

conditioning appli- the annex to this chapter. They were bestRncefrgeratio and air estimates at the time of the study and are not

ance ownership at the time of each study wasestimated in the light of data from industry, necessarily still valid. Modeling included theTypical failure, repair, recharge and replace- producer sector, and the feedstock industriesmTypica ratesre, restabis,rchaged byd dis lacu s to this sector that would be impacted signifi-ment rates were established by discussions catybphso.with local businesses. We modeled, for each Tyb paseout.type of appliance, the numbers and age of The particular difficulties experienced iappliances in each year, adding new appli- India related to the complexity of the indus-

I' .. r ~~~~~~~trial structure in the country and the limitedances in line with production figures, remov-ing appliances from the population based on availability of data on price and demand. Thisaverage life expectancy, and calculating de- ruled out any modeling based on price elastici-

mand for CFCs for recharge using typical ties of demand. Inevitably, we had to makerecharge rates. informed judgments in order to produce

"1order of magnitude" results.V7ery high costs will be associated with "odrfmante"esl.Veary replacmnts ofll refrigerati appcianes wiWe calculated cash flows for four cases: for

early replacement of refrigeration appliances hg-adlwdmn siae ihpaewhen CFCs were no longer available for re- high- and low-demand estilmates with phase-charging. Therefore, we investigated the out as late as possible consistent with the

charging. Therefore, we investigated the Montreal Protocol schedules (allowable) andpossibility of recycling to help meet the ongo-ing demand for recharge refrigerants. We as early as practical (accelerated). For theestimated the volumes of CFCs likely to be high-demand and accelerated phaseout caseavailable for recycling based on the appliances for Group I substances (CFCs), the cash flowsbeing scrapped in urban areas as they reach were as illustrated in Figures 2-3, 24, andthe end of their natural lives and the quantities 2-5.likely to be recoverable from these machines. The rising cash flows in the producer re-Thisdefined supply.Therecovered CFCs were flected the ongoing higher costs of someassumed to be recycled, first to meet recharge substitutes and alternatives. In the user sector,demand for domestic refrigerators and then there is a peak in the cash flows as facilitiesdeman fordomesic rfrigrator andthen are converted to use the substitutes and alter-other appliances. Costs of the recycling opera- atives, th se the higher andtaltertion were estimated from international sources natives, with some ongoing higher costs. Incombined with the structure of the local ser- the consumer sector, there is a surge of costsvicing organization. as appliances are replaced prematurely. The

corresponding (high-demand, acceleratedphaseout) present values are presented inTable 2-2.


We investigated the opportunities to reduce the introduction of recycling to prolong thethe costs in the consumer sector by recycling supply of CFC- 12 for recharge.CFCs. The effect on the economic cost to India(high-demand, accelerated phaseout) is pre- Turkeysented in Table 2-3. The present values for all We found good awareness in industry offour cases modeled are provided in Table 2-4. the need to switch to non-ODSs in the future,

For the high-demand, accelerated phaseout although some companies were waiting for thecase, the breakdown of costs between different Government of Turkey to declare its policysectors is shown in Figure 2-6. We did not runa large number of iterations to find the least- bore theyaact. wnte lrefi geatior,, ~~some companies were already investigatingcost solution, although we established that details about the production of non-CFC prod-early action led to much lower total incremen- ucts (including the manufacture of non-CFCtal costs than delayed action. compressors), and they were very aware of the

The package of policy measures we need to remain competitive, especially in theiradopted to drive this phaseout comprised acombination of market-based and command- insexport markets in Europe. In the aerosoland-control measures, selected for effective- industry market foreslad influncedrthe

nessgivn th chractrisics f te Inian changeover to LPG propellants In many prod-neconomy (ivenclud the wharacridf dstrteIndi ucts. In the foam industry, concerns over costsuesectnors aincudincidhen idelofycorrptriouthed of changeover and lack of experience in the

use of alternative blowing agents were delay-measures included: ing change.

i production quotas on CFCs 11 and 12; ODS use was more limited in Turkey than

* taxes on CFC-l1 and cFC-12 to increase it was in India, and all ODSs were imported,their prices to those of substitutes and mostly from Europe, but also from the Unitedalternatives; States, Brazil and Taiwan. Calculation of

country incremental costs followed the same* financial asslstance to owners of non-swing methodology as applied in India. Country

CFC plants to adapt their units to make incremental costs, expressed as net presentmnore HCFC-22; values, for two cases are set out in Table 2-5.

* incentives to export HCFC-22 to improve Optimum phaseout was defined as phaseoutcapacity utilization; incurring the lowest country incremental cost,

* assistance to aerosol fillers to modify their and allowable phaseout was defined as phase-facilities to use liquefied petroleum gas out occurring as late as possible consistent(LPG); with the Protocol schedules. The net present

values (NPVs) are based on a discount rate of* prevention, through the licensing system, 10 percent real.

the construction of new capacity to build The effect of recycling programs on therefrigerators using CFC-12 as the refriger- early replacement costs is estimated in Tableant; and 2-6. The types of policy measures recom-

* legislation to ban uses in specified sectors. mended for Turkey included:

Overall, we concluded that accelerated * establishing a licensing system for ODS

phaseout significantly reduced the economic users and ODS import monitoring proce-costs to India by reducing the number of dures;appliances that had to be replaced prema- * setting up a task force within the Ministryturely. This cost could be reduced further by of Environment to develop voluntary agree-


ments within industry in order to establish * announcing a phaseout schedule;

specific end-use bans; * establishing a licensing system for ODS

* increasing public awareness through pub- users to restrict distribution and encouragelicity for ODS reduction technologies; environmentally-responsible handling

* introduction of a levy on ODSs to a level practices; andthat makes the substitutes and alternatives * applying a levy on imported ODSs andlook attractive; and setting up an import monitoring procedure

* the eventual banning of the use of ODSs in for ODSs.

specific applications. Zimbabwe

Given the extent to which the market is initiat-ing phaseout, some of these measures may At the time of this study (December 1991),now be superfluous. all ODSs were imported into Zimbabwe, with

Three specific projects for submission for the majority coming from South Africa, Ger-

financial assistance from the Multilateral Fund many, and the United States. Components,were identified. These were all put forward by such as compressors, were also imported.one of the major refrigerator manufacturers in Suppliers seemed to be keeping their custom-the country, eager to work closely with the ers informed about plans for phaseout. MarketWorld Bank team. Interestingly, the other forces, principally the expectation of dwin-refrigerator manufacturers preferred to act dling supplies, were encouraging companiesindependently and did not get very involved in Zimbabwe to plan for phaseout. We werewith the team. There was no attempt to seek told that substitution is commercially viable,out other projects, nor to prioritize them. in the sense that additional costs can be passed

on to customers. There were concerns about

Jordan foreign exchange availability to meet theadditional costs of the non-ODS imports.

Jordan is heavily dependent on imports. All CFCs for recharge have been in short supplyits ODSs are imported, most from Europe and for some time, constrained by foreign ex-some from the United States. All the refrigera- change restrictions. However, we found notor manufacturers import their compressors provision for maintaining a supply post-phase-and only assemble the appliances locally. All out for recharge purposes. We anticipated thatthree companies reported that they would this would mean that the average householdswitch to CFC-free models in line with advice would have to replace its refrigerator aboutfrom their respective suppliers. Some very five years sooner than normal.low-density foams were blown using CFCs, but Government policies to drive ODS phaseoutcompanies will shift to slighter higher densi- did not seem to be necessary. However, weties that do not require CFCs once CFCs are no identified the government role as:longer available. Some aerosol fillers wereusing CFCs as propellants, but the companies * first and foremost, signing the Protocol toconcerned were seeking suitable LPG supplies secure a continuing supply of ODSs intoChange-over was being strongly market-led by Zimbabwe after 1 January 1993, as all thethe influence of developed world suppliers. suppliers are in signatory countries;Policy recommendations to the Government of * reviewing foreign exchange allocations toJordan included: foster the use of substitutes; and

* enhancing public awareness, including * overseeing or coordinating arrangements toemphasis on conservation measures and ensure a continuing CFc-12 supply forsubstitute technologies; recharge purposes.


Given this situation, we did not undertake forces fully taken into account in phaseoutany cost modeling for Zimbabwe. However, planning.we identified that the principal economic coststo Zimbabwe were likely to be Country Incremental Cost Concept

* the additional costs of imported substitute The country incremental cost concept netssubstances and components; and off savings against costs of phaseout. Typi-

* potentially, costs associated with the forced cally, the projects that result in overall savingsearly replacement of refrigeration equip- to an enterprise (such as changing the propel-ment if it fails in service after the time at lants in aerosols) occur first, because it is inwhich there is no CFC-12 available for the company's financial interests. Increas-recharge. ingly, such phaseout has already taken place

before a Country Program is prepared. There-

Lessons Learned fore, future country incremental costs over-state total costs, as the savings have already

Market-Driven Phaseout been realized and are not included.Country incremental costs are highly uncer-

There IS a strong market influence on tain, because:phaseout, especially in countries that importtheir ODSs or components for use with ODSs * They are based on long-range projectionsfrom developed countries. The market influ- that will always be subject to considerableence, often mirroring developed country uncertainty; andphaseout dates, can leave the Article 5 country * Unit incremental costs can only be esti-governments with few, if any, choices. The mated.government role can be limited in these cases; The country incremental cost conceptpolicies such as taxes and regulations may notbe needed to drive change. However, training * is useful for testing alternative possibleand awareness activities driven by govemment phaseout timings and determining least-may continue to be important, especially in cost strategies for a country overall. How-some specialized applications where direct ever, where market forces are causingsubstitution is not the only option and supplier phaseout, the government may not needinformation may be insufficient or biased. such detailed projections to drive policies,Country Programs need to reflect the more since it is not making choices;limited government role in these cases and * does not contribute to selection of suitablemay therefore be much simpler than expected. policy measures, although it may guide the

ODS producer countries are more insulated timing of their introduction;from the influence of market forces fromdeveloped country producers, although the * may be useful for sequencing investments;influence still persists where there are joint andventures or technology agreements with devel- * does not identify project costs for the pur-oped country producers. A Country Program poses of assistance from the Multilateralbased on the country incremental cost concept Fund.may still be relevant in these situations, wheregovernment policies need to be selected to Confidentiality and Transparencydrive phaseout. Some industrial companies in Article 5

Experience suggests that ODS producer and countries are initiating phaseout to maintain ornonproducer countries need to be treated improve their competitive position. Thesedifferently, and the strong influence of market companies rightly feel that their phaseout


plans, including timetable and costs, are Dominance of Early Replacement Costscommercially confidential in a competitive The country incremental costs of ODS

market. A national phaseout plan needs to phaseout are domlnated by the costs of earlyrespect this confidentiality, replacement of appliances when CFCs are no

In these reports, we have made every effort longer available forlrecharging. This cost isto ensure that economic cost calculations are sensitive to the size of the population of CFC-

transparent by including details of all cost and bgaoother assumptions, and describing the ratio- baerfigatsanarcodinrsnueother asspion ad m descin. bing the a when CFC supply ceases. The size of theprblem binIndithe modreling.sTi was not a populatlon will generally be lower as soon asperobearle inagIndia wher aepase wae ratn a substitute appliances are introduced into thevery .earl osage .and esi ates werdra market. This argues for phaseout as soon as

from itae possible. The supply of CFCs for recharge canthere are potential conflicts where companies pole chave specific plans that they do not wish to be recyclngCe, for byasourcing dreoping sbt

released ito the pubic domain.recycling CFCs, or by sourcing drop-in substi-tutes. The costs of this need to be balanced

Project Costs against the potential early replacement costs.Governments may be best placed to initiate

In practice, projects seem to be put forward action to create mechanisms to ensure a con-for funding from the Montreal Fund independ- tinuing supply of recharge material compatibleent of influence on the country incremental with cFC-12 appliances.costs. The estimates of country incrementalcosts become redundant, and detailed model- Project Prioritiesing does not seem to be appropriate in such Projects submitted for assistance from thecases, especially since country incremental Mc.osts are not equal to the sum of all Identified MutlerlFnmabehoewseic-proectcosts.are noteqAlto, the sue nof adenied, imental costs are easiest to identify rather thanthe projetacosts Aso, w vnoy probmse, in strategically important leadership projects, ortheos preparciation of cutr program, con- those that will help the early replacement inscou reasconcilirategioneofswhat.shouldbefundd the consumer sector. This suggests that someto least-cost strategies. v

Even at the project level, true incremental way of prioritizing is needed for the mostcosts can be difficult to identify. In practical effective use of international funds for phase-

terms, it is most realistic to limit applications out.for funding to capital costs, and exclude on- Conclusionsgoing costs. This seems to be the practice.Any project funding in a competitive economy The conditions set out for the Multilateralis effectively a subsidy to the companies F treceiving the assistance and may distort the asssof the Montreal Protocol for providing

market. ~~~~~~~~assistance for ODS phaseout to developingmarket. countries have proved difficult to apply inIn many small countries, projects tend to be prcieConyicemtacssaenta

smal n ers o te otl ivetmntre practice. Country Incremental costs are not asmall in terms of the total investment re-quired. Respov mvery helpful measure. They are extremely

quird. Rsponive eansof povidng asis- difficult to estimate and do not pinpoint thetance in such cases are needed, with the mini-

mum of administrative overhead. It maybe areas in the economy that are most critical formum o admnistativ overead.It my be effecting phaseout, nor where the costs will

more efficient for implementing agencies to fall.deal directly with companies in disbursing N isn tassistance rather than through governments.N ditnioismeinheersf

assistc rreference of the Multilateral Fund between thedifferent situations in ODS producer and


nonproducer countries. The type of assistance The Producer Sectorthat is required by chemical producers, with In estimating the costs in the producercomplex plants and capacity utilization impli- sector, we have modeled the capital and oper-cations, is quite different from that needed in ating cash flows for ODS plants, substitute

equipment manufacturing sectors. plants or substitute imports, and for significantExperience shows that the additional costs feedstock industries. In this section we also

per unit of consumer product are a very small show the assumed capacity increments andproportion of the total cost of the product and e u the consumer can bar te irse project Where a feedstock industry is costed sepa-

rately, the cost of the raw material it suppliesto the oDs-substitute plants is excluded from

Annex 2-1: Unit Cost Assumptions the operating costs of these subsequent plants

The following is an extract from the Phase to avoid double counting. For both HFC-134aII report prepared by Touche Ross in 1990 for and HCFC-123, we have assumed that whenthe Government of India on "Reducing the production capacity is constructed in India, it

Consumption of Ozone-Depleting Substances will comprise 5,000 tons per annum units.This is yer small In comparison with the

in India." It describes unit cost assumptionsand is included with this chapter to illustrate scale of plants which are being designed andefforts to make the calculations transparent. built in developed countries. Therefore, it may

prove to be difficult to acquire designs and

The Costs of Substitution: Introduction specifications for small units. If small unitsare built, they may not enjoy the economies of

As described in the section above entitled scale available to producers in developed"Methodology Used to Calculate Country countries. The transition from existing sub-Incremental Costs," the costing methodology stances to substitutes may therefore requireused takes account of costs on a cash flow the import of substitutes for a period of time.basis. For the producer sector, the full capital All costs are expressed in 1990 prices.and operating cost cash flows are calculated Capital costs include plants and machinery,first to meet continuing unconstrained demand land and buildings, utilities, and commission-for ODSs ("base case") and then to meet the ing. As is normal practice in India, land pricessame level of demand with substitutes. The take into account infrastructure costs. For newdifference gives the incremental cost of com- processes, technology transfer costs are identi-pliance to India. In the user and consumer fied separately. Fixed operating costs includesectors, the incremental costs (conversion of labor, overhead, rates and insurance, andexisting facilities and additional operating maintenance, but not depreciation, which is acosts) are identified directly. This section sets noncash item, nor return on capital, which isout the unit costs for each sector. allowed for when the cash flows are subse-

Costs are expressed in rupees and pounds quently discounted to 1990.sterling. An exchange rate of Rs 28=fl is usedand, in every conversion, the resulting CFC-11/12 and HCFC-22 Manufactureamounts are rounded. In presenting the rupeeamounts, we have used the Indian conventionInmeigbscaedad,ewC-

ramounts as f ows: 11/12 and HCFC-22 plants will be required.

We have assumed that it would be similar to

Rs 1 crore = Rs 10 million; and the recently installed swing plants with capac-ity increments of 5,000 tons, although this is

Rs 1 lakh = Rs 100,000. relatively small by international industry


standards and may fail to generate available ment. The small amount of existing carboneconomies of scale. The costs included for tetrachloride production from a rayon plant issuch plants reflect the cost structure in the assumed to cease during the phaseout period.Indian industry today and are set out in Table We have assumed that carbon tetrachloride2-7. produced as a by-product from chloroform

With phaseout of ODSs, two of the five production will be incinerated at high temper-CFC-l 1/12 production plants can progressively atures (in excess of 1, 100 degrees Celsius) indecrease CFC- 1/12 production and switch to a purpose-built furnace, with all the requisitefull HCFC-22 production; these are the newest effluent treatment facilities, including wet gasplants, and are swing plants. One of the two scrubbing. We have researched cost estimatesold cFc-11/12 only plants is very small and in the United Kingdom, where highly effectivewill become redundant. The other two plants modern incinerators are already operatedcan be converted to full HCFC-22 production by commercially, and the costs assumed are setadjusting the operating conditions. Additional out in Table 2-9.HCFC-22 capacity will be installed as required,with capacity increments and costs the same HFC-134aas the modem swing plants given in Table The principal substitute for CFC-12 as a2-7. refrigerant is assumed to be HFC-134a. We

Chloromethanes have assumed that demand for HFc- 134a willbe met initially by imports and subsequently

Additional chloromethane capacity would from domestic production once demand isbe needed to meet the base case projections. sufficient to justify a local plant. Cash flowSuch a new plant is assumed to be similar to projections therefore include the cost of im-the most recent existing plant, using the chlo- ports until these are supplanted by a localrination of methane process. It will have the product. The estimated long-run sales price ofcapacity to produce a total of 6,000 tons per HFC-I34a is taken to be five times the currentannum of chloroform and carbon tetrachlo- cFc- 12 sales price (both net of any import orride. Cost estimates are based on information other duties imposed by the government ofcollected in the United Kingdom and are set India), as is anticipated by companies cur-out in Table 2-8. rently developing HFC-134a.

With phaseout of ODSs, chloromethane The capital investment and operating costscapacity will continue to be required to supply of local plants are included as incurred. Thethe chloroform feedstock for HCFC-22 produc- feedstocks required for HFC- 1 34a productiontion. Fewer new plants will be required than in are still commercially secret, but could bethe base case, but no plant closures will be ethylene and HF acid, both of which arenecessary. The demand for carbon tetrachlo- readily available in India. Capacity incrementsride will decline both as a feedstock for of 5,000 tons have been assumed, as thisCFC-l1/12 and as it is phased out for use in seems appropriate to the scale of demand incleaning and as a solvent. India. However, this may not achieve the full

This will pose a problem for chloromethane benefit from economies of scale.producers, as it is not possible to eliminate Again, there is a wide range of views as tocarbon tetrachloride production and still what is likely to be the minimum economiccontinue to produce chloroform. This is true size of a plant, ranging from 5,000 or 10,000even though the proportion of carbon tetra- tons up to 50,000 tons. At such an early stagechloride in the product mix can be reduced in the development process for HFC-134a, onlyfrom the current 15 percent to about 8 percent order of magnitude estimates of costs areor less with minimal additional capital invest- available. The capital costs are expected to be


roughly double the costs of CFC plants in the We have included a one-time payment of 25United Kingdom, allowing for some progress percent of the total capital cost for illustrativedown the learning curve before a plant is set purposes. See Table 2-11.up in India in 2005. It is implicit in the analy-sis that the technology will be made available Hydrofluoric Acid and Fluorsparto India. The terms for such a transfer are More hydrofluoric (HF) acid plants wouldunknown as yet, but could take the form of a be required to meet base case demand. It isone-time payment or a royalty payment. We assumed that this will be similar to existinghave included a one-time payment of 25 apercent of the total capital cost for illustrative plants. Cost estimates are based on the costpurposes. See Table 2-10. structure in the Indian industry today and are

set out in Table 2-12.

HCFC-123 With phaseout, Hi,- will continue to be afeedstock for HCFC-22, and will also be re-

We have assumed that HCFC- 123 (a substi- quired for the production of HFC-134a andtute for some CFC-l I applications) will be HCFC- 123, which require more HF per productimported initially and then manufactured than the CFCs they will replace. Overall, thelocally once demand is sufficient to justify an tonnage of HF required will be a little less thanindigenous plant. Cash flow projections there- that for base case demand, since the substitutefore include the cost of imports until these are production volumes are projected to be lesssupplanted by local production. The estimated than those of CFCs.sales price of HFC-134a is taken to be 2.5 Fluorspar will also be required in slightlytimes the current CFC- 11 sales price (both net reduced quantities compared with the baseof any import or other duties imposed by the case. Although there are plans to source fluor-government of India), as is anticipated by spar indigenously, we do not at this stage havecompanies currently developing HCFC-123. details of the likely scale and price of domes-

The capital investment and operating costs tic production. Cash flows therefore assumeof local plants are included as incurred. The an uninterrupted supply of imported fluorsparfeedstocks for HCFC-123 are ethylene, chlo- at today's landed prices.rine, and HF acid, all available locally inIndia. Capacity increments of 5,000 tons have Methyl Chloroformbeen assumed, as seems appropriate to the New methyl chloroform plants will bescale of demand in India. However, as with needed to meet base case demand. It is as-HFC-134a, this may not fully benefit from sumed that these will be part of vinyl chlorideeconomies of scale. compe like thee eisting pant. Costs have

Again, there is a wide range of views as to complexes, like the existing plant. Costs havewhat is likely to be the minimum economic been estimated on this basis, based on U.K.size of plant, ranging from 5,000 or 10,000 experience.tons up to 50,000 tons. At such an early stage production will have to be phased out com-in the development process for HCFC-123, pletely. Most of the vinyl chloride complexonly order of magnitude estimates of costs are wltl, however, be unaffected with the bulk ofavailable. wits output, cniungfte potyetie into

The capital costs are expected to be roughly its output, continuing to be polymerized intodouble the costs of CFC plants in the United polyvinylchloride (PVC). See Table 2-13.Kingdom, allowing for progress down thelearning curve before a plant is set up in India CFC 113in 2005. It is implicit in the analysis that the All CFC-1 13 demand is currently met fromtechnology will be made available to India. imports, costed at the landed price, excluding


all import and other duties collected by the * trichloroethylene replacing methyl chloro-government of India. As has been true from form and carbon tetrachloride in metal

early 1991, demand will be supplied from cleaning;

local production. * carbon dioxide and dry chemical powdersIn addition to the two new CFC- 113 produc- replacing halons;

tion facilities about to come on stream, further organic solvents replacing carbon tetrachlo-plants will be required to meet base case ride when used as a solvent.demand. It is assumed that this will be a stand-alone plant, and not an add-on to cFc-1 1/12 Our cost estimates for these substitutes andcapacity. Cost estimates are based on both the tons of substitute required per ton of ODSU.K. and Indian industry experience. The replaced are given in Table 2-16.1,000 ton capacity increments are small by In the user sector (manufacture of productsinternational standards, but match the pattern using or containing ODSs) only incrementalof demand build-up projected for India. See costs are identified. These include capitalTable 2-14. The CFC-1 13 plants will become costs to modify or convert existing facilities,redundant as use of these ODSs ceases and they incremental capital (over and above thatwill close down. required before substitution) to build new

facilities, or incremental operating costs.

Halons These are discussed below. All costs are

All halons used in India today are imported. quoted in 1990 prices.

We have costed them at the landed price, Domestic refrigeration

excluding all import and other duties collectedby the government of India. As from the Costs of switching to water blowing (usingbeginning of 1992, demand will be supplied a water/CFC- I I mixture) for the cabinet insu-from local production. lating foam are assumed to be negligible,

In addition to the new halon production based on U.K. experience to date. Beyondfacilities about to come on stream, further this, two types of change are required in exist-plants will be required to meet base case ing domestic refrigerator manufacturing facili-demand. Cost estimates are based on Indian ties to produce HFC- I 34a/HCFC- 123 refrigera-

industry experience and use the same 1,000 tors:

ton capacity increments as the plants that is * minor changes to the integrated foam blow-currently planned. See Table 2-15. The halon ing facilities to work with a wa-plants will become redundant and will close as ter/HCFC- 123 mixture;use of these ODSs ceases.

* some modifications to the production lines

Not-in-Kind and Other Substitutes (principally retooling) to accommodateredesigned compressors, evaporators and

Some substitutes are costed in the model by condensers that will optimize power con-the inclusion of the substitute purchase price sumption with HFC-134a refrigerant.(net of all Indian taxes) rather than by thecapital and operating costs of a manufacturing in Dusons th friator manufact

plant Theyapper in he prduce sectr lnEurope suggest that, for a typical manufac-plant. They appear in the producer sector turing unit in India capable of producing

figures (Figure 2-3) for consistency. The 500 ,000 refrigerators per annum, the conver-following substitutes are treated in this way: . c

sion costs will be In the order of Rs 1.3 crore* LPG replacing CFCs as aerosol propellants; (£0.5 million). In addition, a technology trans-

* surfactants replacing CFC- 113 and methyl fer fee may be involved, for which we have

chloroform in electronics cleaning;


included Rs 1.5 crore (£0.5 million) for each CFC-12 room air conditioner compressor500,000 refrigerator unit. manufacturers should not incur any significant

We do not expect there to be any incremen- incremental costs in converting to HCFC-22tal capital costs associated with new capacity. units. There will be no conversion costs orThat is, new capacity will cost the same to change in assembly costs in either the orga-build whether it is designed to produce CFC or nized or unorganized sectors, since theHFC-134a/HCFC-123 refrigerators. HCFC-22 units are essentially similar to the

An increase in operating costs in foam cFc-12 units they replace.blowing (excluding the extra cost of substitute Most reciprocating central air conditionersHCFC-123, which is included in the producer are currently built using HCFC-22 compres-sector cash flows) may result from the use of sors; no changes will be required. The remain-different polyols and more process control. der of production which still has CFC- 12However, any such increase is expected to be compressors will require redesigned compres-very small. sors, and therefore the production lines may

European refrigerator manufacturers are not need some retooling. We have assumed thatanticipating any incremental costs of manufac- this will cost in the order of Rs 50 lakhsturing HFC-134a compressors, evaporators, or (£0.2 million) per 1,000 compressor unit.condensers (excluding the extra cost of substi- There are not likely to be any incrementaltuting HFC-134a as refrigerant, which has been operating costs, over and above the extra costincluded in our costing model in the producer of the substitute refrigerant.sector cash flows). Some centrifugal central air conditioner

compressors manufactured today are fullyOther refrigeration compatible with HCFC-123 and will require no

Water an otlcoes,ieadfurther modifications. For the remainder,WVater and bottle coolers, ice candy ma-chines, deep freezers, and cold stores will all some conversion costs will be incurred inrequire redesigned compressors, evaporators, retooling the production line. These are likelyand condensers to operate with HFC- 1 34a. The to be the same for reciprocating compressormanufacturing facilities tend to be smaller for lines, although the manufacturing unit size isthese systems. The cost of retooling compres- smaller (only about 1,000 each per annum).

sor-lines.is estimated to b Rs I crore (LOA There are not likely to be any incrementalsor lineS IS estimated to be Rs 1 crore (£04 y4million) per 100,000 units of annual produc- operating costs, over and above the extra costtion capacity for all such lines, other than of the substitute refrigerant.those for ice candy machine compressors, for Car air conditioners will also require somewhich we have estimated the cost to be redesign to operate with HFC- 1 34a, and there-Rs 10 lakhs (£35,700) per 5,000 units of fore there will be conversion costs incurred by

annual production capacity. The assembly the manufacturers to accommodate this inoperations will be unaffected. their facilities. Indications are that these will

There are not expected to be any incremen- be roughly Rs 40 lakhs (£l43,000) per 50,000tal capital costs associated with new capacity, production line. There are not likely to be anynor any incremental operating costs for com- incremental operating costs, over and abovepressor manufacture or for assembly of the the extra cost of the substitute refrigerant.finished goods. Foam blowing

Air conditioning Other than the changed costs of the blowing

Most room air conditioners are already agents themselves, which are costed in themade with HCFC-22 compressors. The few producer sector, conversion costs to modify


the blowing nozzles and flow lines will be required, since the CFC- 11 currently acts as anegligible. solvent, but this is likely to be a one-time

development cost, not incurred in India.Aerosols

We have assumed that aerosol cans will Electronics vapor degreasingbegin using LPG as a propellant, except for The changeover from CFC- 113 and methylinhalants, which will use HFC-134a when it chloroform to aqueous cleaning systems willbecomes available, and that bottles will begin generally require slightly increased batchusing mechanical pumps. sizes, heaters to warm the water baths and for

LPG to be used as propellants in aerosols drying, and the installation of effluent disposalmay be purchased in and used from cylinders, facilities. Based on survey work in the Unitedor bought in bulk quantities and stored in a Kingdom, the capital costs involved are esti-sphere or bullet. In the former case, the cost mated to be around Rs 3 lakhs (£11,000) for aper kilogram of the LPG is significantly higher small batch operation (equivalent to an exist-than a bulk purchase, while in the latter a ing unit with an annual consumption of 0.6sphere or bullet must be purchased. The deci- tons of cFc-1 13 or 1 of methyl chloroform).sion will hinge around the volumes involved. Offsetting this, operating costs (excluding theWe have assumed bulk purchases and the use cost of the surfactants, which are costed in theof storage bullets. producer sector) will decrease by about

LPG also introduces fire risks to the aerosol Rs 24,000 (f850) per ton of ODS solventfillers. Therefore, safety precautions (includ- replaced. Incremental capital costs for newing a ventilation system and a gas detection facilities are around Rs 2 lakhs (£7,000) forsystem) will have to be introduced at each small batch operations.filling line; relocation to more suitable pre-mises may also be advisable. Costs for each of Metal vapor degreasingthese elements, based on experience in the Research in the United Kingdom showsUnited Kingdom, are estimated to be in the that when methyl chloroform is substitutedorder of Rs 10 lakhs (£35,700) per filling line. with trichloroethylene, there is no loss ofThe investment will not be economic for very cleaning effectiveness, change in processingsmall filling units; we have assumed that those time, changes to process facilities, or incre-filling fewer than one nmllion cans per annum mental operating costs.are more likely to switch to the mechanicalpump alternative. A similar incremental in- Metal dip cleaningvestment is assumed for new capacity.

There do not appear to be any incremental Similarly, the use of trichloroethylene inoperating costs when filling aerosol cans using place of carbon tetrachloride incurs no penal-LPG as the propellant. The conversion of bottle ties. There is no loss of effectiveness andaerosols to mechanical pumps does not require cleaning times will be very similar.any significant changes to the filling stations.However, the pump valves cost 50 percent Incremental Cost Assumptions for themore (an additional Rs I or 3.5p each) com- Consumer Sectorpared to aerosol valves. We have assumed that Only owners of refrigeration and air condi-these will continue to be imported. tioning equipment incur incremental costs in

Early indications are that there will be no the consumer sector, as we have defined it.costs associated with switching filling stations These are costs that have not already beento HFC- 1 34a propellant for inhaled aerosols. included in the producer and user sector andThere may be some product reformulation fall into two broad categories:


* incremental running costs, which would be placed. The cost to the consumer of forcedincurred by all owners of non-ODS equip- replacement is therefore represented by thement; timing difference between the actual early

* forced replacement costs, which would be replacement and the normal replacement. Theincurred only by those owners whose CFC cost of a new refrigerator is included in thebased equipment fails in service but cannot cash flow at the time of early replacement, andbe recharged once CFCs are no longer avail- an equal and offsetting credit is given at theable. projected time of normal replacement.

This additional demand for replacement

Running costs refrigerators will increase the required supplyof refrigerators, refrigerant HFC- 1 34a, and

The latest evidence from development in foam blowing agent HCFC- 123 and will there-the use of HFC-134a and HCFC-123 as refriger- fore impact the producer and user sectors. Forants indicates that there will be no energy simplicity of presentation, however, we havepenalty with their use. Although HFC- 134a is not separately priced these effects but havethermodynamically less energy efficient than included instead the full cost of a replacementthe existing cFc- 12, the combination of rede- refrigerator in the consumer sector cash flows.signed compressors and new, less viscous Similarly, the costs of early replacement oflubricants is considered by informed sources compressors in air conditioners and otherlikely to result in energy savings. We have refrigeration units are attributed in full to theconservatively assumed that, overall, the consumer sector.substitution is energy neutral and that there Because of the scale of these early replace-will be no cost or savings for the consumer, ment costs, we have evaluated how they canand there are indications that there could be a be reduced by the use of recovery and recy-net energy benefit. cling, or by stockpiling, which can extend the

period of time for which CFCs are available forForced replacements recharging. The possible structure and cost of

Costs to some consumers may arise from a recycling industry to meet the demand forthe forced replacement of domestic refrigera- recharging substances is described below.tors, air conditioner compressors and icecandy machines, and other refrigeration com- Recyclingpressors when no CFCs are available for re- A recycling industry, subdivided into rela-charging. This could be a particular problem tively self-contained parts, may emerge inin India where consumers continue to use India in response to economic opportunities inequipment for much longer than in developed the market place as follows.countries, prolonging the useful life by repair Mobile air conditioners. Collection of usedand recharging. CFC- 12 will take place at a service point

We have calculated the number of units that when the vehicle is taken for compressorare likely to be forced to be replaced prema- repair and recharge. The used cFc-12 canturely, when CFC availability ceases, based on be reprocessed on site and reused for re-an estimate of the number of refrigerators that charging other mobile air conditioners. Thewould have been recharged in any year were amount of cFc-12 reclaimed will not beCFCs still available. In the forecast population sufficient to supply all recharge demandof refrigerators, the age of each refrigerator at from mobile units, since leakage occursthis time is known. Therefore, we know when during normal use. Also, there tend to bethis refrigerator would normally have reached losses from mobile air conditioners, andthe end of its useful life and have been re- these emissions will not be available for


recovery. Each service point will require a be acids, moisture, carbon particles, andreclaim unit and two full-time, trained metal swarf that can be removed by filter-operators. ing. Badly contaminated CFC will be col-

Central air conditioners. Large volumes of lected and returned to a CFC producer forCFC-12 are involved in reciprocating units, reprocessing.and of CFc-1 1 in centrifugal systems. A In this latter case we have worked on thereclaim machine will be taken to the plants basis that collection is only likely to occur into extract used CFC and to clean it up, al- the ten major urban conurbations in India. Welowing recharge to take place on-site. Some have assumed that there will be a number ofadditional CFC may be required for topping distribution centers in each of these cities,up. When reciprocating systems reach the each with a reclaim unit and two trainedend of their normal life, the CFC-12 can be operators. The additional cost of collectionextracted and added into a bank of recycled from the individual refrigerator and delivery tomaterial for use in other applications, such the distribution center is likely to be veryas domestic refrigerators. small, since the engineer will visit to diagnose

* Refrigeration, including domestic refriger- the problem, whether recharging is required orators. When a service engineer goes to not.repair and recharge a domestic refrigerator, The same type of reclaim unit can be usedhe will take with him a simple portable in each sector and costs about Rs 126,000collection device to collect the tiny volume (£4,500) each. Running costs of the machinesof used CFC- 12 from the refrigeration sys- are negligible, although two operators may betem before he starts. He will take this back required for each. Portable collection devicesto a distribution center. Here the collected cost about Rs 280 (£10) each.CFC-12 will be placed in batches with other Our estimate of the total costs of establish-used CFC- 12 and reprocessed on the spot, ing and operating a significant recyclingsince most of the impurities in the CFCs industry in India are set out in Table 2-17. Allcollected from domestic refrigerators will costs are in 1990 prices.

42 OzONE LAYER PROTECTION: COUNTRY INCREMENTAL COSIS

Table 2-1: Uses of ODSs in India and Possible Substitutes, 1990

Substitute Assumed earliestSector Current use assumed availability in India

RefrigerationDomestic refrigerant CFC 12 HFc 134a 1998Industrial refrigerant Ammonia Ammonia NowWater coolers, bottle and walk-in coolers, ice CFC 12 HFC 134a 1998

candy machines, cold storage unit refriger-ants

Fridne!i sulatinz foam CFC 11 HCFC 123 _ 1998

Air conditioningRoom HCFC 22, cFc 12 HCFC 22 NowPackaged HCFC 22 HCFC 22 NowCentral

reciprocating HCFC 22, cpc 12 HCFC 22 Nowcentrifugal CFC 11 HCPC 123 1998

Mobilecars CFC 12 HFC 134a 1998buses, vans cFc 12 HFC 134a 1998trains cFc 12 HCFC 22 Now

ships- HCFC 22 HCFC 22 Now

FoamsFlexible polyurethane CFC 11 HCFC 123 1998Rigid polyurethane CFC 11 HCFC 123 1998Phenoic CFC 11, 113 HCFC 123 1998Poly cCFc 12 HFC 134a 1998.P-oJ-s-rene _______-_- __-c- C 12 _ HCFC 22 Now ------

AerosolsInhaled phamaceuticals CFC 11, 12 HFC 134a 2005

Bottle aerosols CFc 11, 12 Mechanical pumps NowCan aerosols cFCI11 12 LPG - Now

CleaningVapor degreasing

electronics CFC 113 Aqueous 1995computer hard disc memory CFC 113 None knownelectronics Methyl chlorofom Aqueous 1995metal cleaning Methyl chlorofom Trichloroethylane Now

M_ta_dipclean.ng-_... _Carbon tetrachloride Trichloroethylane NowFireflghtingPortable units Halon 1211 None knownFixed flooding units ----------------- _Halon 1301 --------- None known --------------------

Process solventsPesticides, pharmraceutical, industrial, paint Carbon tetrachloride Other organic sol- Now

manufacture vents


Table 2-2: Producer Sector Annual Incremental Cash Flows (high-demand, accelerated phaseout), Group I (CFCs)

Rs crore £ millionTotal economic cost (presentvalue, 1990) 8% 12% 8% 12%

Producer sector 360 170 130 60

User sector 120 80 40 30

Consumer sector 440 250 160 90(no recharging after 2010)

TOTAL 920 500 330 180

Table 2-3: India: User Sector Annual Incremental Cash Flows (high-demand)for Group I (CFCs)

Economic cost, present value (1990)

Rs crore million

Recycling strategy 8% 12% 8% 12%

No recharging after 2010 920 500 330 180

Limited recharging 870 460 310 160

Significant recharging 630 330 220 120

Full recharging 490 254 170 90

High-demand scenario used.

Table 2-4: India: Consumer Sector Annual Incremental Cash Flows (high-demand) for Group I (CFCs)

Economic cost. present value (1990)

Rs crore £ millionScenario/alternative 8% 12% 8% 12%

High-demand/acceleratedphaseout 920 500 330 180

High-demand/allowablephaseout 2,110 1,065 755 375

Low-demand/acceleratedphaseout 735 400 265 145

Low-demand/allowable phase-out 1,860 935 660 325

"No recharging after 2010" strategy assumed.


Table 2-5:Turkey: Breakdown of Net Present Valuesfor Each Sector (high-demand) for Group I (CFCs)


Optimum phaseout Allowable phaseoutUS$ million US$ million

Incremental capital costs 2 1

Incremental operating costs 69 31

Forced early replacement costs (no 56 200recycling)

TOTAL 127 232

Table 2-6. Turkey. Phaseout Costs


Optimum phaseout Allowable phaseoutUS$ million US$ million

Early replacement costs with:

* no recycling 56 200

* limited recycling 34 117

* substantial recycling 10 71

Table 2-7. Unit Costs of CFC-11/12 and HCFC-22 Swing Plants

Capacity increments 5,000 tons

Capital costs Rs 11 crore £4 million

Fixed operating costs per annum Rs 2 crore £0.75 million

Variable operating costs per ton* Rs 3,600 £130

* Excluding raw materials (HF, chloroform, and carbon tetrachloride) for which costs are included in the cost estimatesfor the feedstock industry. One ton of CFC-1 I requires 0.17 tons HF and 1.24 tons of carbon tetrachloride. One ton of CFC-12 requires 0.38 tons HF and 1.43 tons of crc. One ton of HCFC-22 requires 0.62 tons of HF and 1.54 tons of chloroform.

Table 2-8: Unit Costs of Chloromethane Plants




Variable operating costs per ton Rs 9,900 £350


Table 2-9: Unit Costs of Chlorinated Hydrocarbons Furnace



Technology transfer 25 percent of capital costs payable in foreign currency


Variable operating costs per ton* Rs 1.5 lakh £5,400

* Excluding HF raw material for which costs are included in our cost estimates for the feedstock industry. We haveassumed that one ton of HFC 134a will require 0.66 ton of HF.

Table 2-10: Unit Costs of HFC-134a Plants



Technology transfer 25 percent of capital costs payable in foreign currency


Variable operating costs per ton* Rs 40,000 £1,430

Excluding HF raw material for which costs are included in our cost estimates for the feedstock industry. We haveassumed that one of HCFC-123 will require 0.66 tons HF.

Table 2-11: Unit Costs of HCFC-123 Plants



Technology transfer 25 percent of capital costs in foreign currency


Variable operating costs per ton* Rs 40,000 £1,430

* Excluding HF raw material for which costs are included in our cost estimates for the feedstock industry. We haveassumed that one ton of HCFC-123 will require 0.66 tons HF, and that each ton of HF requires 2.2 tons of fluorspar asfeedstock.


Table 2-12. Unit Costs of Hydrofluoric Acid Plants


Capital costs Rs 9 crore £3.2 million



Table 2-13: Unit Costs of Methyl Chloroform Plants



Fixed operating costs per annum Rs 0.7 crore £.25 million


Table 2-14: Unit Costs of cFc-113 Plants



Fixed operating costs per annum Rs I crore £0.35 million

Variable operating costs per ton* Rs 17,300 £610* Excluding hydrofluoric acid raw material for which costs are included in the cost estimates for the feedstock industry.One ton of CFC 113 requires 0.36 tons of HF acid.

Table 2-15: Unit Costs of Halon Plants






Table 2-16: Incremental Cost Assumptions for the User Sector

Tons of substitute/Rs/ton of substitute Liton of substitute ton of ODS replaced

LPG 16,800 600 1.00for 11/12

Surfactants 84,000 3,000 1.56 for 1131.00 for MC

Trichloroethylene 17,500 625 1.38 for MC

1.00 for crc

Carbon dioxide 38,250 1,370 3.00 for halon

Dry chemical powders 50,000 1,785 Not known

Organic solvents 13,800 490 1.00 for CTC

Table 2-17: Costs of Recycling Industry

Capital215 reclaim units Rs 2.8 crore £1 million100,000 portable collection units Rs 2.8 crore £1 millionOperating

Labor Rs 3.4 crore £1.2 million

Portable collection units renewal Rs 2.8 crore £1 millionCosts to buy in contaminated CFC 12 (20 percent of Rs 10,320/ton £370/ton

virgin CFC price)


Figure 2-1: Example of Phaseout for Domestic Refrigerators (high-demand, accelerated phaseout)

1.61.7I.e1.5

14 watar2 ubfAd o 14 blowinig l3Ulfitdo

1.3 l oanm1.2

O.1 replae 12

31 0.90.6

0.7 0.6

0.40.3 - rehag0.2-coi rd0.1 by1 pl

1990 1993 13 2010

* Now 12 - Recharge 12 -* Foam I11 Total

Figure 2-2: Elements of the Model

p.d.- ODS ~- -- 1h

Pd-,~~~~ -rwh n u r mlios

hfl. -1 nch or -rcUy


Figure 2-3: Producer Sector Annual Incremental Cash Flows (high-demand, accelerated phaseout)for Group I (CFCs)

220

200- 70

160 0

Al~~~~~~~~~~6

140 50

120--4

0~~~~~~~~~~~~~~~~~~~~4

1 100 0

b 3

8s0 - 0

60 - 20

40

0 _ I L X. 10

0 0

-20

140 gn90 192 194 1619Q98 2doo 2do2 2004 2d06 2008 2010 2d12 2014

Y.

Figure 2-4.' User Sector Annual Incremental Cash Flows (high-demand, accelerated phaseout) forGroup I (CFCs)

40

35

12

30-

10

25

20 -

15

'0 4

5 2

0 990 1992 1694 1496 1 68 2dO 20 2 20O4 20O6 2d 6 2d10 2 132 14T


Figure 2-5: Consumer Sector Annual Incremental Cash Flows (high-demand, accelerated phaseout)for Group I (CFCs).

1.413 0.51.2

0.4

b 0

0.8 0.3 00.7

0

0~~~~~~~~~~~~5 ~~~~~~~~~~0.2 0.4

0.3 0.1

0.2

0.1

0 ~~~~~~~~~~~~~~~~~~~~01990 1992 1994 199 Z1-998200220 2 06 2008 2010 2012 201

Figure 2-6: Breakdown o?f Net Present Values for Each Sector (high-demand, accelerated phaseout)fior Group I (CFCs)

1000~~~~~~~2

PROWkCER USER CON0UMtERSECTOR SECTOR SECTOft

:8500

13~~~~~~:~.~ i

40tt-0

154t() _ _ __ _ _ _ __ _ _ _

3

Evaluation of Incremental Cost Methodologiesfor Phasing Out Ozone-Depleting Substances:

Case Study of Egypt

Charles B. Catanach

ACCURATE ESTIMATES OF COUNTRY INCREMENTAL COSTS forphasing out ozone-depleting substances (ODSs) are critical in thedevelopment and evaluation of alternate national policies andODS phaseout strategies for developing countries. The policiesand strategies selected in the estimating process form the basisfor a Country Program to phase out ODSs. This Country Programbecomes a commitment by government upon which financialgrants are made by the Multilateral Fund under the MontrealProtocol to assist the phaseout. Egypt was one of the firstcountries in which such cost data were evaluated with theassistance of the World Bank (July-August 1991). Subse-quently, the Country Program was prepared, reviewed, andsubmitted by the government of Egypt to the Executive Com-mittee of the Multilateral Fund. It was subsequently approved inOctober 1992.

51


This chapter details how the incremental costs materials, equipment, and technologies werefor Egypt were estimated and used in the estimated, and the direct incremental costs ofprocess of preparing its Country Program. the phaseout were compiled. The forecast was

then used to estimate inventories of ODSs

Calculating using equipment (such as refrigerators) whichCountry Incremental Cost would be subject to early obsolescence for

each case studied. The consumption forecastThe methodology for calculating country was the key element in the analysis. It was

incremental costs drew on experience by the developed by adjusting the unconstrainedauthor in forecasting the impact of the Mon- demand for ODSs for all significant elements,treal Protocol on the international markets for forcing conversion to non-ODSs for eachchlorofluorocarbons (CFCs) and on early strategy evaluated.World Bank documents which outlined theguidelines and important considerations for Projection of Demandestimating incremental costs (King andMunasinghe 1991). In the case of Egypt, it is Unconstrained demand is the baseJine forimportant to distinguish between financial strategy development and is defined as thecosts to enterprises, incremental costs which demand forecast for chlorofluorocarbonsmay be eligible for grant financing through the (CFCs) and other ODSs as if the ozone issue hadMultilateral Fund, and country incremental not evolved into the Montreal Protocol andcosts, which are used to evaluate country local business conditions remained steadyprogram alternatives. Financial costs are through 2010. The baseline provided discretedetermined at the enterprise level and include data against which to forecast the impact ofprofitability elements. Multilateral Fund grant the Protocol-related constraints, conservationfinancing is based on the "Indicative List of or substitution, recovery or recycling programCategories of Incremental Costs" included in consequences, and other conditions whichthe appendices to the Montreal Protocol and might increase or decrease the actual demandthe policies of the Executive Conmmittee of the for CFCs or its substitutes.Multilateral Fund. Country incremental costs Actual demand is defined as:include estimated Multilateral Fund grant Actual = UC - Conserve - NIK + New Uses

financing plus the economic costs to a country where: Actual = Actual ODS Demand (without Proto-resulting from early obsolescence of equip- col constraints)ment and technologies based on the continued UC = Unconstrained ODS Demanduse of ODSs. Specifically excluded from coun- Conserve = Material savings via recovery ortry incremental costs are customs, tariffs, and recycle, othersduties on the imported goods needed to ac- NIK = Not-in-kind substitutes for ODS

New Uses = New applications requiring ODScomplish the phaseout. substitutes

Note: All units must be in common ODS unitsor equivalents so they can be converted to

The overall process for estimating costs for actual quantities via appropriate factors relatedthe Egypt Country Program required the to physical properties or actual experience.preparation of a spreadsheet ODS consumption Actual consumption is the actual demand asforecast which could be modified as necessary constrained by the conditions imposed by theto reflect various Montreal Protocol (MP) Protocol and multiplied by the percentage ofcompliance and country strategies. With this forecast consumption not replaced by substituteforecast in hand, the needs for replacement products (for instance, the hydrofluorocarbon

HFC-134a substituting for CFC-12).

3. Incremental Cost Methodologies for Phasing Out ODSs: Egypt 53

The demand for ozone-depleting substances A large domestic market, a diversified(ODSs), chlorofluorocarbons (CFCs) in particu- industrial base, ideal agroclimatic conditions,lar, has traditionally been related to basic a favorable location and a large, skilled laborsocial and economic conditions which impact force give Egypt strong potential to achievethe quality of life. In general, the state of the noninflationary growth and a viable balance ofeconomy has a direct bearing on gross na- payments. Future growth depends primarily ontional product (GNP), which in turn relates to the implementation of meaningful economicthe consumption per capita of ODS. As GNP reforns without jeopardizing social and politi-increases, disposable income expended on cal stability. If accomplished, the demand foritems other than food and housing will also domestic goods such as home refrigeratorsincrease, and consumers will tend to purchase and furniture is forecast to grow again at 5 toincreasing numbers of items related to their 10 percent per year; such growth will directlystandard of living, such as refrigerators, air impact the demand for ODSs and alternatives.conditioners, and miscellaneous household In the Egypt case study, a conservative as-luxury items. Marketing activities by multina- sumption was made of a growth rate increas-tional producers of CFCs have recognized this ing to 3 percent per year by 1996. The uncon-relationship and commonly based growth strained demand for CFC products was as-assumptions for CFCs on GNP growth. Glob- sumed to increase at a rate of 2.7 percent perally, a logarithmic relationship has also been annum faster than the GNP per capita growthdeveloped to relate CFC consumption to GNP. rate while also increasing directly with theA characteristic of most countries has been population (currently 2.6 percent per annumthat consumption of CFCs exceeds GNP growth growth)by at least two percentage points. The highest Sectoral differences also affect the forecastgrowth is observed in countries moving from growth rate, particularly when emerging ODS

poverty to economies in which there is some consuming industries, such as electronics, arediscretionary income. introduced into a country. In Egypt, this was

Egypt's economy has varied widely in the not a factor and the 2.7 percent growth factorpast 10 years. In the decade from 1974-1985 was used in all sectors. Spreadsheet methodol-Egypt's economy grew at 8 percent per year. ogy, however, can readily adjust to reflectDespite this growth, a widening gap between sectoral differences.foreign exchange needs and earnings led to a Table 3-1 illustrates growth assumptionslarge foreign debt situation (greater than 100 used to project growth of unconstrained de-percent of GNP), and ultimately to debt re- mand for CFCs and other ODSs from actual ODS

scheduling in 1987. In response to this situa- consumption data.tion, the government initiated a series ofpolicy changes which reduced price controls, Specifying Montreal Protocolinitiated reforms in public industries, and Constraintsencouraged foreign investment. The periodiclack of foreign exchange distorted regular The Montreal Protocol imposes directimports of raw materials (such as CFCs). In constraints on domestic consumption of CFCs

turn, this led to distortion of production when and other ODSs. In recent years, these con-foreign exchange was short and to stockpiling straints have been modified significantly.of unused materials when foreign exchange Spreadsheet interpretation of these constraintswas available. Overall, the available historical requires the following data:statistics for CFC use in Egypt are difficult to the base year for estimation of constraints;interpret with confidence. *tebs erfretmto fcntans

* the percentage of the base year consump-tion allowable;


* ozone-depleting potential (ODP) weighting * 15 percent to fill and service large unitsfactors for each controlled substance. (such as chillers)

The spreadsheet objective was to calculate The primary replacement for CFC- 12 inthe allowable ODS demand so that it could be most applications was assumed to be HFC-

compared with forecast ODS demand to deter- 1 34a, the alternative of choice by the majoritymine if forecast ODS demand had to be con- of global refrigerator manufacturers. Replace-strained as a result of the Montreal Protocol. ment of CFC- 1I in commercial large refrigera-While developed countries have specific base tion units and chillers was assumed to be viayears upon which to key their phaseout pro- retrofit with the hydrochlorofluorocarbongrams (such as 1986 for CFCs), Article 5 Para- HCFC-123. New chillers were assumed to begraph 1 developing countries have a 10-year designed to use HFC-134a. It was also assumedgrace period for phaseout. Under the condi- that HFC-134a and HCFC-123 availabilitytions of the Protocol for Article 5 Paragraph 1 would not constrain their application as alter-countries, base years will be no earlier than natives beyond 1994.1996. Consequently, the ODS quantities upon The low cost of cFcs has discouragedwhich to base future Protocol restrictions in conservation practices during servicing ofArticle 5 Paragraph 1 countries cannot be refrigeration equipment. While it was assumeddetermined until 1997. This could impose a that improved servicing practices will developrather interesting constraint on countries that as CFC prices increase in the future, CFC- 12move aggressively to substitute alternatives service requirements were assumed to remainfor ODSs, as in doing so they may reduce the constant at an average of 50 grams per yearquantity of ODSs otherwise allowable under CFC-12 per active refrigerator. A more basicthe Protocol. assumption was that refrigerators in Egypt

required a full replacement charge of CFC-12

Defining Technical Options once every five years. Recovery or recyclingand Assumptions activities related to the refrigeration sector

were forecast to directly reduce demand for

Technical options for implementation of CFCs, depending upon the strategy underthe ODS phaseout have been thoroughly evalu- analysis with the spreadsheet cost model.

ated and published in the reports of the Tech-nical Options Committee for the Montreal FoamsProtocol. On a global basis, these preferred Demand for CFCs in the foam sector derivestechnologies are similar for each ODS end use. from the ideal characteristics provided by CFCsThe costs of each replacement technology and as blowing agents for insulating foams inthe timing of its introduction are critical fac- appliances, housing, and miscellaneous auto-

tors~ ~ ~ ~ ~ ~ ~~~~~phacs impacting inreeda costs.no atotors impacting incremental costs. motive applications and for flexible foams

used for cushioning in mattresses and furni-ture. The effectiveness of insulating foams

Six-hundred-and-eighty tons of CFCs were relates to the thermal conductivity of theused as cooling fluids in Egypt during 1990 blowing agent and the cell size of the finished(Arif 1991). Usage was as follows: foam. To date, no alternative has demon-

* 20 percent to fill new refrigerators during strated comparable performance to the CFCmanufacture. blowing agents. Replacing up to 50 percent of

the CFC- II requirement with water has pro-* 65 percent to service refrigerators and duced satisfactory insulation in a large portion

related equipment. of the refrigerators made in Europe. In the

future, the short-term replacements are as-


sumed to be HCFCs (for instance, HcFc-141b) cFcphaseout rates fromaccelerated to allowable.

and hydrocarbons (for instance, cyclopentane). For the Egypt case study, nine cases were

In the long term, application of non-ODS analyzed under the Protocol constraints in

technology is considered likely. existence in 1992. Three phaseout timing

For flexible foams, CFCs are currently being alternatives were analyzed at three different

replaced with methylene chloride. By 1994 in economic growth scenarios.

Egypt, it is assumed that all cushioning needs The phaseout timing alternatives were:

will be replaced with substitutes that are not in 1. Allowable phaseout. The transition tokind or with methylene chloride blown foams. non-ODS substitutes is not encouraged in

Solvents Egypt until forced under the terms of the

Montreal Protocol (that is, complete phase-

The demand for ODS solvents in Egypt is out by 2006).

low in comparison to the demand for CFCs. A 2. Accelerated phaseout. The transition tosimple assumption is that all ODS solvents ODS substitutes is accomplished at a rate

would be phased out via new cleaning or no- consistent with the availability of technol-

clean processes or through redesign of compo- ogy and is fully accomplished by 1998.

nents requiring cleaning.3. Optimum phaseout. The rate transition to

Halons ODS substitutes results in the minimumcountry incremental cost. The details of this

As with solvents, the domestic demand for case were developed by iteration of the

halons in Egypt is small. Again, it is assumed basic cost model. However, it was eventu-

that a phaseout will be feasible within the ally determined that the timing for the

constraints of the Protocol. It is assumed that optimum alternative lagged behind that of

critical needs will be met from stockpiles of the accelerated alternative by approxi-

halon. The stockpile would be formed from mately one to three years.

halon recovered from existing installed equip-

ment, retrofit with another extinguishant such

as carbon dioxide (CO 2 ) or water. * High-global economic factors favor a GDP

per capita growth in Egypt of better than 4

Outlining Alternative Strategies percent per year.

* Reference-most likely economic situation.There are an infinite number of alternative Therowth ratelo gronatio prouct

strategies and scenarios which could be con-

sidered as likely for an ODS phaseout. ODS (GNP) per capita is given in Table 3-1.users serve a wide variety of applications. * Low-zero economic growth. The growth

Some ODS users can benefit from immediate rate of GNP per capita is given in Table 3-1.

replacement of CFCs with less costly substi- The growth assumptions were used primar-

tutes, while others would be severely impacted ily to develop ranges for country costs calcu-

by an unduly rapid CFC phaseout. Country lated under the different phaseout timing

economic conditions directly impact ODS alternatives, thereby providing an estimate of

demand, technology development is uncertain, sensitivity for each alternative to economic

and Protocol constraints are likely to be tight- factors.

ened. In developing the strategy framework

for Egypt, only the most likely high, baseline,

and low economic growth rates were consid-

ered. Also considered under the Protocol were


Recognizing Incremental Costs * public awareness programs to reduce ODS

The "Indicative List of Incremental Costs" use;included in Appendix vm of the Protocol * institutional capacity building;document was the primary reference for * improved monitoring of ODS imports;determining eligible incremental costs. Ingeneral, such costs are borne at the enterprise * safe replacement of CFCs with liquefiedlevel as a direct result of Protocol phaseout petroleum gas (LPG) in aerosols;requirements. Only those activities underway * recovery or recycling equipment and pro-in 1989, when the Protocol went into force, grams;were considered to be eligible for funding by * reclamation of recovered CFCs; andthe Protocol Multilateral Fund. The forecastsof unconstrained and actual ODS demand * establishment of a halon bank.established useful benchmarks and providedindicators for areas in which incremental costs Estimating Obsolescence Costswere likely to be incurred. Only those costitems judged to be significant were included Costs associated with premature replace-in the final analysis. ment of equipment designed for use of ODS

Net incremental operational costs which are represents a significant cost burden whichexpected to qualify for grant financing through must be borne by ODS equipment owners. Itthe Multilateral Fund were also included in also represents a direct cost burden to thethe analysis. Such costs are net of savings and economy of a country and is therefore consid-are generally related to the following items: ered to be an incremental country cost in the

Egypt case study. Whereas the incremental* cos onn-DSramteias;capital and operating costs for the transition to

* labor and supervision; and non-ODS for Egypt are estimated to be about

* plant efficiency. $50 million, the early obsolescence costscould approach $200 million, depending on

Significant projects which were considered the phaseout strategy followed by Egypt. Theto qualify for funding through the Protocol major item subject to early obsolescence is theMultilateral fund and which were incorporated household refrigerator. If it is assumed thatin the Egypt case study analysis were: Egypt continues to import or manufacture CFC

v retrofit of chillers and large refrigeration refrigerators until forced to stop under theunits; terms of the Montreal Protocol, essentially all

* retooling of domestic refrigerator manufac- refrigerators in Egypt would become obsoleteand would require replacement within five

ture; years (that is, whenever they fail).

* redesign and retooling of water cooler Figures 3-1 and 3-2 compare refrigeratormanufacture; populations for two strategies and clearly

* facilities to make HFC- 134a compatible illustrate the effects of an early phaseout oncompressors the refrigerators subject to early obsolescence

plant improvements for safe useofmethyin 2010.* plant improvements for safe use of methy- Estimation of obsolescence costs for

lene chloride; Egypt's refrigerator fleet required estimates of

* high pressure foam machines for rigid the following:foams; * total number of refrigerators in service in

* technical assistance and retraining; the base year;

* annual production of CFC refrigerators;


* annual production of non-ODS refrigerators; * The current constraints imposed by theand Protocol are known. The amount of ODS

* average useful life of refrigerators. allowable under the Protocol for each sub-stance is calculated for comparison with the

Using spreadsheet methodology, the num- ODS demand forecast. The allowable con-bers and average ages of the fleets of CFC and sumption is then calculated to estimate thenon-ODS refrigerators were calculated for each ODS consumption forecast (for example, ifyear through 2010 under each phaseout strat- the allowable consumption exceeds theegy. It was then assumed that all such CFC calculated demand, the consumption isrefrigerators would become obsolete in 20 10 taken to be equal to demand; otherwise,and the increment cost for obsolescence could consumption is taken to be the allowablebe calculated based on the average crc refrig- level).erator age. It was also assumed that suchrefrigerators had depreciated on a straight-line Assumptions are made concerning thebasis over 20 years. alternative strategies and scenarios.

Model Description Calculation FlowForecast demand is projected from a histor-

General ical base (pre-1991) of data using growth

The model used to simulate costs for the scenarios and other key assumptions appropri-Egypt case study was developed as a single ate to the case being analyzed. The populationspreadsheet. While it might have been feasible of various types of refrigerators and the needto use the same modeling concept with multi- for recovery or recycling machines to serviceple spreadsheets, laptop memory and spread- installed equipment is estimated. Demand issheet recalculation speed appeared to impose adjusted to reflect the ODS involved. Next,severe limitations. The simplicity of a single forecast demand is compared to that allowablespreadsheet offered more options for iteration under the Protocol and the forecast is con-of a wide variety of strategies prior to full strained, where necessary, to estimate ODSdefinition of the final cases included in the consumption. Capital and operational costcountry study. It was also possible to quickly estimates were prepared off-line to reflect theeliminate insignificant considerations by magnitude and timing of the interventionssimple what-if analyses. needed to accommodate the forecast of ODS

consumption. Finally, all costs are accumu-

Structure of the Model lated year by year, and the totals are dis-counted at 10 percent to the base year to

* Key assumptions for the strategy under estimate the net present value of the cost tostudy are GNP and population growth, rate Egypt for the phaseout strategy under analysis.of transition to alternative substances orprocesses, total phaseout investment and Summary of Countryoperational costs, and miscellaneous con- Incremental Costsstants. The impact of recovery or recyclingoperations on CFC supplies and on invest- Cost to Egyptment requirements are also estimated.

The methodology facilitated tabulation of* A forecast of ODS demand is then prepared eahcscopntyarbya.Tre

using these assumptions. Estimates of total cos onental cs were deeODS eman an ODPconribuionare ade categories of incremental costs were devel-ODS demand and Pr cntri inaed oped for each case studied-capital, opera-

tional, and early obsolescence. A set of three

58 OZONE LAYER PROTECTION: COUNTRW INCREMENTAL COSTS

potential country growth rate scenafios was Policy Instruments Usedanalyzed for each phaseout alternative (see The following policy instruments wereTable 3-1). The high and low growth ratecases were then used to establish a probable proposed for the Egypt Country Program:range for incremental country cost for each * A panel of industry and government lead-strategy (see Table 3-2). ers, referred to as the "Ozone Panel," to

provide a consultative role in the formula-Sensitivities tion and promulgation of specific policies

to be incorporated into the National OzoneThe spreadsheet technique provides a Plc.Tercmeddplce nld

convenient process for many types of "what- voluty agreements with sindusyif' sensitivity analyses. Most analyses are estabi specends bans cndustry tostrongly influenced by one or more key as- wish shecoerl oectvebs of itheny ~~~~~~~with the overall objectives of the ODSsumptions, and the degrees of freedom are phaseout and to recommend timing oflarge. In the Egypt study, as in other similarcountry studies, however, the effect of refrig- merits.erator obsolescence overwhelmed most othercontributors to country phaseout costs. There- * A licensing system for ODS users whichfore, assumptions impacting obsolescence encourages environmentally responsiblerequired careful scrutiny before inclusion in ODS handling practices and restricts distri-the final model. Growth scenarios, as illus- bution to users qualified through approvedtrated in the above-mentioned figure, had only training programs.minor impact on total phaseout costs and had * A formal ODS import monitoring system asno significant impact on the ordering of the required to meet Protocol conditions.alternative strategies. Consequently, the * Selective price increases through importgrowth scenarios were used mainly to develop taxes to minimize the financial cost for

substituting alternative materials or pro-

Recommending Policies cesses for the ODSs used in key industries.

Overall Strategy Impact on Country Incremental Costs

Based on the analysis of costs, the optimum Recommended policies in the Egypt Coun-overall strategy for ODS phaseout in Egypt was try Program were primarily designed to facili-one which followed the phaseout in developed tate an effective phaseout of ODS and will onlycountries by one to two years but where cost- impact country costs by delaying the phaseouteffective phaseouts were accomplished imme- if not adopted promptly and effectively. Moni-diately (for instance, in the case of aerosol toring ODS users and licensing of ODS handlersconversion to LPG propellant). There are, are examples where action is needed to assurehowever, practical limitations on how fast that country phaseout goals are not contra-such phaseouts can be accomplished. Limita- vened.tions relate to the availability of technologyand licenses, the lead time for construction of Lessons on Methodologynew facilities, as well as the availability of Recommendations for Future Modelsforeign exchange and financial resourcesprovided through the Multilateral Fund. In This case study, and other similar studies,addition, flexibility must be maintained revealed that the most cost-effective strategythroughout the phaseout to meet unforeseen involves phaseout of ODS on a schedule whichdevelopments. is significantly more accelerated than that


required under the Protocol. The basic ques- suppliers or distributors to develop a tabula-tion is, therefore, how best to accomplish this tion of prices of key materials free-on-boardaccelerated phaseout given the existing infra- key global ports (such as Rotterdam, Mar-structure and available financial resources. seilles, Houston, San Francisco, Tokyo, andThe development of models has a role in Singapore). Freight costs would then be esti-identifying the country ODS consumption mated independent of product type for use inpatterns and future needs and the elements of specific country studies and projects for con-potential costs. The need for optimization sideration by the Multilateral Fund.does not appear to exist, as the most practicalcost-effective accelerated phaseout is the Unit Abatement Costsoptimum one. Models are also of value as a The methodology for developing bench-mechanism to mobilize local involvement in

the dei .nkn thog idniyn dat mark unit abatement costs needs modificationto reflect capacity effects. It would be desir-

needs, structuring "what-if' strategies, and able to have unit abasement costs which candeveloping ownership over the final results. be compared directly between similar projects

Other lessons apply to the methodology and and between countries so that modeling ofsuggest that for the best results, the model capital cost estimates can take full advantageshould be logical, easy to apply in the field, of experiences on projects with technologicaland easy to learn. Internally, macros should beused to develop all results to assure consis- sugges b ent adjust t

suggested procedure would be to adjust thetency and to avoid mistakes. incremental capital cost in a unit abasement

Spreadsheet models often reflect the per-sonality of the model-builder and can contain pcost baseline for a similar operation by the 0.6hidden information or assumptions. There is pef,bl capacity d ed thealso a common tendency to overstate theaccuracy of such models even though they tion.contain hundreds of interactive assumptions. Example:A good way to overcome such potential prob-lems is to identify the key variables in an Data

iterative process and then to analyze why theyare important and, if necessary, to make ap- Actual capacity of plant = 350,000 appli-propriate adjustments to the model structure. ance units per year

In the final analysis, the spreadsheet should Unit abasement cost for project to convertonly guide the study of the problem and sup- xyz plant to non-CFC foams = $27.32/kg.port the conclusions and recommendation. ODS saved

Areas for Improving Cost "Standard" plant capacity = 1,000,000

Calculations appliance units per yearBaseline unit abasement cost for similar

Unit Prices for ODS and Alternatives foam project in a "standard" plant =

Basic assumptions of unit prices for materi- $13.75/kg ODS saved

als imported into developing countries have Calculation

major impacts on the estimated operationalcosts for an ODS phaseout. A methodology is Estimated unit abasement cost for xyzneeded for estimating future prices for ODSs project = (1,000,000/350,000)06 x 13.75 =

and replacement materials. One way to do this $25.8 1/kg ODS savedwould be to enlist the support of multinational


Project Cost Effectiveness or country differences yielded significanteconomies to the XYZ project.

The project costs and ODS savings for theXYZ project are comparable ($27.32 vs. Bibliography$25.81) to those for a "Standard" project eventhough the unit abasement cost is almost two Arif, S. K., and Charles B. Catanach. 1991.times higher than the baseline unit abasement Assessment of Ozone-Depleting Substancescost ($13.75). in Egypt, Yugoslavia, Turkey and Tunisia,

A constraint would be that the adjustment March, World Bank, Washington, D.C.should only be made on the basis of discrete Egypt-Country Program for Phaseout ofoperating units, such as manufacturing lines. Ozone Depleting Substances. Approved atIf the xYz project was to convert two similar the Eighth Meeting of the Executive Com-manufacturing lines, the calculation would be mittee of the Multilateral Fund, Octoberas follows: 1992. UNEP/OzL.Pro/ExCom/8/14.

Estimated unit abasement cost for XYZ King, Kenneth, and Mohan Munasinghe.project = (1,00O,OOO/175.000)06 x 13.75 = 1991. "Incremental Costs of Phasing Out$39.12/kg ODS saved Ozone Depleting Substances." Environ-

ment Working Paper No. 47, World Bank,The conclusion would then be that for some Washington, D.C.

undefined reason, either the xYz project waspossibly underestimated, the scope was notdirectly comparable to the "standard" project,


Table 3-1: Growth Assumptionsfior Reference Demand

Growth rate/yr 87 88 89 90 91 92 93 94 95 96 97

Population growth 2.6 2 6 2.6 2.6 2.6 2.6 2.6 2.6 2.6 2.6 2.6GNP per capita 5.0 0.0 -6.6 -5.5 -2.3 -2.3 -1.0 00 1.0 30 3.0Sectoral factor 2.7 2 7 2.7 2.7 2.7 2.7 2.7 2.7 2.7 2.7 2.7Net growth mte for 10.3 5 3 -1.3 -0.2 3.0 3.0 4.3 5.3 6.3 8.3 8.3

ODSs

Table 3-2: Accelerated, Optimum, and Allowable Phaseout

Accelerated phaseout Optimurn phaseour Allowable phaseout(1993-98) (1994-98) (2000-10)

US$ tnillions US$ millions US$ millions

Net incremental capitalcosts 38 37 16

Net incremental opera-tional & administrativecosts 20-30 19-29 6-11

Early replacement costsfor refrigerators 4 6 71-119

TOTAL 62-72 62-72 93-146

a.'

Rsrigerwot (millions)

Rdtgeratos (MillK)

E w NS y eNwrS U Mf<ees4St0j e>a e%wywe

s __ ww w w w w ~~~ ~~~ ~~~ ~~~~~~~~ -ww.Sv ...... _ , ,, X X ..0 ._ 5~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.. . .. . .. .

9 _ , W .__ j jKW .__j ............................ _.S aL_ > ss >NDS < MA _~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. . . . . . . . . . . . . . . . .

b _ , _ ,. _ . . ... a -- _ ' ' ' ' ' ' ' >~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~......

_ M . .. o _ n j _ ... ... .................... _0.NE.<e O. <<ONO) \::::r rY q -O~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.. ... .. ..

Z 0~~~

gl0g k~~~~~~~~~~~~~~

~~~~~~~~~~~~~~~~~ N ~ ~ ~~ ~~ ~~~~~~~~~~~~~~~~~~~~~~~~c

4

Evaluating the Incremental Costsof Complying with the Montreal Protocol:

Discussion of a FrameworkDeveloped for Thailand

Vikram Widge, Mark Radka, and Thomas Dillon

Background

THIS CHAPTER IS BASED ON WORK PERFORMED in conducting thecountry study on the phaseout of ozone-depleting substances(ODSS) in Thailand. It presents an analysis of ODS consumptionin Thailand and recommended certain national policies andactions to meet the global ODS phaseout.

The study was conducted for the Department of Industrial Works of Thailand's Ministry ofIndustry by a three-part team consisting of ICF Incorporated, SLAMTEC Intemational, Ltd., andthe Thailand Development Research Institute Foundation, and was completed in September1992.

63


Thailand's consumption of ODSs is not large Montreal Protocol (the "baseline").by world standards, but until recently it had Incremental costs can be either imposed orbeen growing rapidly. On an ozone-depleting voluntary. Imposed costs include increases inpotential (oDP)-weighted basis, imports in chemical prices brought about by Protocol-1991 were 10,044 weighted tons, indicating a induced supply shortages; because Thailandper capita consumption of 0.16 kg per person does not produce ODS itself, these costs arefor chlorofluorocarbons (CFCs) and halons and faced whether or not Thailand complies with0.02 kg per person for l,l,l-trichloroethane the Protocol. Voluntary costs include the cost(methyl chloroform, or MCF) and carbon of new equipment needed to produce goodstetrachloride (CTC). Major use sectors include without using ODSs; these costs are incurredsolvent cleaning in the electronics and metals only if Thailand complies with the Protocol.industries, domestic and commercial In ICF Incorporated's original version ofrefrigeration (both original manufacture and coscOM, the chemical cost component ofservicing), vehicle and building air control costs was not adjusted to account forconditioning, production of rigid polyurethane the projected rise in the price of ODSs or theinsulating foam, and fire protection. fall in alternative chemical prices. Instead, we

used constant ODs and substitute prices, partlyCalculating Country because price projections were tenuous, atIncremental Costs best, at the time the study was conducted. ODS

prices are rising because the phaseout isIntroduction to the Methodology making the chemicals scarce, not because the

resource or social costs of production areIncremental costs for the phaseout of ODSs c g although s rcity til fre

were calculated using a simulation modelcalled COSCOM (the Country Study Cost consumers to pay more for ODS-basedModel) that was developed for the Thailand products, extra costs paid by consumers areCountry Study.' We chose to use a simulation transfer payments and not true social costs in

model because sImulatioii provides a tool by a global sense. Money is being transferredmodel because simudelasystion pro are tool bfrom consumers to producers or importers,

which t modelsystem that re not providing them with supranormal profits, butnecessarily amenable to a mathematical .-representation and, therefore, to standard no additional resources are being expended in

producing ODSs.7 Incremental costs are theoptimization techniques like linearproptimriziong Simulatechniq s lpproprikeliear costs incurred over the baseline; in the

progammig. Smulaion as apropiatefor baseline, there is no phaseout, and ODS pricesthis analysis because the optimization of thephaseout plans and costs had to be performed do not rise.

Thai decisionmakers, however, weigh theovaraterimutiple end useswithdiverse present and projected future prices of ODS andcharacteristics.

The incremental cost of phasing out ODSs in non-ODS technologies when makingconversion decisions. We believe that theThailand is the difference between costs

incurred under a chosen phaseout strategy and most accurate way to calculate incrementalcosts is to determine the sequence of

costs corresponding to the series of actions ithat would have occurred in the absence of the

2. An ODS tax can be designed to redistribute transfer1. Although some of the substitutes of CFCS, halons, payments from ODS producers and importers to theand 1,1,,-trichloroethylene are also ozone depleters, in government. In essence, this tax will transfer to thethis chapter we use ODS to refer only to the chemicals government any supranormal profits resulting frombeing phased out per the London Amendments to the price increases that are caused by productionMontreal Protocol. restrictions.

4. Evaluating the Incremental Costs of Complying with the Montreal Protocol: Thailand 65

phaseout strategies using unit abatement costs growth rates for ODS demand. This addedthat have variable ODS and substitute prices, complexity was warranted because: (1) theand compute the costs relative to a non- premature retirement or retrofitting of theseProtocol baseline. This approach was made items-a critical component of overallexplicit in later versions of our model, and costs-is very sensitive to projections of totalwas accounted for in a rough way in the stock in the final years of the phaseout; (2)reference strategy used in the Thai study. MACs and domestic refrigerators are likely to

The principal inputs to COSCOM were (1) account for a significant portion of the growthprojections of unconstrained ODS demand, (2) in total ODS use; and (3) reducing ODS use inphaseout schedules with interim targets set by MACs and domestic refrigerators will accountthe Montreal Protocol, and (3) the costs and for a substantial portion of the costs of thecharacteristics control options and substitute phaseout.technologies. Together these inputs For export sectors, two sets of growth ratesdetermined the initial scenario, the set of were used to project consumption. Growth inexogenous factors that affected the analysis exports of electronics was assumed to be 10but over which Thai decisionmakers were percent per year for the first three years of theexpected to have little control. The principal projection (based on historical trends), fallingoutputs were the sequence of actions required to 5 percent per year thereafter. Growth into meet phaseout targets and the incremental other exports sectors was assumed to be 5cost of these actions. The costs were percent per year through 1995, falling to 3computed following guidelines put forth by percent per year thereafter. Growth wasthe World Bank for the Executive Committee expected to stabilize at the lower rates in aof the Multilateral Fund. The following few years as Thai export industries reach theirsections describe in greater detail the model's potential in world markets.inputs, the alternative phaseout strategies In the demand forecasts for the refrigerationmodeled, and the results of the analysis. and air conditioning sector, manufacturing

demand was separated from servicingUnconstrained Demand demand. These were tracked separately in the

Projections of unconstrained demand for simulation because real-world controls affect,, ,. , , ................. .them differentl . For example, as increasingCFCs, halons, and MCF In the air conditioning, y g

refrigeration, solvent cleaning, aerosol, foam, numbers of newly installed MACs are charged' . ' ' . ' ~~~with HFC-134a, the demand for cFc-12 atand fire protection sectors were the startin wihHC14,tedmn o F-2a

point fore prothenaysi.rectionssecto e fr etac installation will dwindle quickly in the MACspoint for the analysis. Projections for each sector. The demand for CFC-12 at service,sector were based on 1991 ODP-weighted ODS ser, depend or stock of exice,consumption data and on growth rates derived however, depends on the stock of existingfronsumpeonoc data fndongrom th Thaie durive. MACs and will decline gradually, disappearing

from conoic dta fom Tai idusties. only after the last vehicles with cFc-12 MACDemand projections were truncated in 2015, othe end year for the analysis. systems are retrofitted or retired. For this

For most domestic sectors, ODS con- reason, it was important to maintain thesumption was projected to grow at the same distinction between manufacturing and chargerate as projections of gross domestic product demand, versus servicing demand in the(GDP) made by the Bank of Thailand (5.7 to model.8.6 percent per year). Mobile air conditioners Phaseout Schedules(MACs) and domestic refrigerators werenotable exceptions. For these sectors, pooled Also figuring in the analysis was thecross section and time-series statistical Montreal Protocol-imposed phaseoutregression analysis was used to estimate schedule, which expresses interim and final


consumption ceilings as percentage reductions * Early retirement controls. Early retirementfrom the consumption levels in a base year. controls cause existing equipment to beThe Thailand analysis used the pre- retired prematurely and replaced with newCopenhagen Montreal Protocol control equipment using alternative chemicals.schedule for CFCs, halons, and MCF. Because Early retirement controls only apply toCOSCOM was not designed to evaluate the equipment with a servicing demand. Earlyimpact of a hydrochlorofluorocarbon (HCFC) retirements typically occur in end usesphaseout, no HCFC phaseout schedule was where a retrofit is technically impossible ordefined. costs more than the value of the equipment.

Phaseout plans for those strategies modeled Different types of controls were allowed towere required to meet the 2010 to 2015 phase- interact in a number of ways. For example, itout, as well as interim caps and other wac t recycling always egins itconstraints, such as the ban on exports to an end use at least as early as the beginning ofdeveloped countries of products containing conversion to alternative chemicals. As aODSs imposed during the 10-year grace period result, implementation of substitution controlsfor meeting Protocol targets. triggers recycling if recycling was not already

Control Options occurring in the end use. Recycling has theeffect of decreasing the demand for virgin

The last set of model inputs was the ODSs and obviating the need for other controls.controls available to reduce ODS consumption Conversely, early retirement and retrofitin each sector. Controls were based on the controls decrease the remaining stock ofauthors' appraisal of the alternatives and equipment being serviced, and diminishes thecontrol technologies likely to be available in reductions achievable through recycling.each sector during the period of the analysis. The following market characteristics wereControls used in the study were of four types: specified for each control in the model:

* Substitution controls. When substitution * Date of earliest availability: the year incontrols are implemented, new equipment which the alternative was expected tois manufactured with alternative chemicals become commercially available. In noinstead of ODSs (for instance, new strategy modeled could a control berefrigerators made with HFC- 134a instead implemented before its date of earliestof CFC-12; foam blown with HcFC-141b availability.instead of CFC-l 1). * Eventual market penetration: the

* Recovery controls. Recovery controls percentage of the unconstrained ODSreduce ODS use through recovery or consumption in the sector that therecycling of ODSs at service or disposal. alternative was expected to displace once itRecovery controls only apply to equipment reached its full market potential. The sumwith a continuing demand for ODS at of the maximum market penetrations for allservice. Recovery controls reduce ODS use, of the substitution controls in a sector wasbut do not eliminate it. set equal to 100 percent so that a phaseout

* Retrofit controls. Retrofit controls replace could be achieved.ODSs with alternative chemicals in existing * Years to reach maximum marketequipment. (that is, a CFC- 1 chiller penetration: the number of years it wouldconverted to use HCFC-123). Like recovery take for a control, once it was introduced,controls, retrofit controls only apply to to reach its maximum possible marketequipment with a servicing demand for penetration. This parameter reflected theODSs.


fact that the transition to a substitute is * Accelerated. The accelerated strategynever instantaneous. assumed that all substitution and recycling

Per kilogram cost to industry: the incre- controls were implemiiented as soon as they

mental cost per kilogram of avoided ODS became conmercially available, withoutuse that industry would incur by regard to cost. If the remaining ODS

demand for ser-vicino existing r-efi-igerationimplementing the control. This included I e I riprojected changes in capital, operating, and and air conditioning equipment was high

chemical costs. enough that phaseout targets werc violated,chemical costs.retrofits and carly retiremiienits were

* Per kilogram cost to consumers: the implemented. from least expensive to mostincremental cost per kilogram of avoided expensive, until comlpliance is reached eachODS use that consumers would incur. This year. This strategy woulld resUlt in theincluded changes in the energy efficiency fastest ODS phascout.of new equipment, changes in the costs ofmaintenance and service, and a portion of thall ala. Th ullowabe streg ansumethe retooling costs in some sectors. at Thaiad wo tathesiium

action required to meet year ly phaseout* Premature retirement and retrofit costs: the targets. In each year in the simulatioln,

incremental cost per kilogram of avoided controls were implemented to meetODS use expected to be incurred as a result immediate phaseout targets with theof retrofits or early retirements. This was minimum marginal cost but without regardeither the cost of the retrofit or the cost of to possible future costs. In contrast to thehaving to pay for new equipment prior to accelerated strategy, no action was takenthe normal replacement schedule on a per that was not dir-ectly forced by thekilogram of ODS avoided basis. constraints of the Montreal Protocol.

Technical data on control options were Controls were impleinenited in order ofobtained from ICF Incorporated's technical increasing expense and were dclayed for asexperts, United Nations Environment long as possible, heing implemented onlyProgramme technology assessments, surveys when needed to meet the phaseoutof manufacturers, and trade literature. In total, schedule. By delaying implementation ofthe model contained information on 41 substitution cotntrols, expensive retrofittingalternatives in seven ODS-using sectors. and early retirement were required after theAlthough this was fewer than the number of phaseouLt date.alternatives actually or potentially available, * Forward-lookinig. The forward-lookingparticularly in sectors like solvent cleaning, strategy was a variation of the allowablemany alternatives share enough cost strategy, in which the future costs ofcharacteristics that the modeled alternatives actions were takenr into accounit by theserved as proxies for related alternatives with model in selecting the controls. As in theno appreciable loss of accuracy in a country- allowable strategy, most controls werelevel analysis. delayed for as long as possible while

meeting the constraints imposed by thePhaseout Strategies Examined Protocol. The forward-looking strategy,

Given the input data described in the however, recognized that decisions in thepreceding sections, COSCOM was used to refi-igeration and air conditioning sectorcalculate total incremental costs for four would consider the trade-off betweenphaseout strategies: accelerated, allowable, delayed implementation of substitutionforward-looking, and reference. controls and increased early retirements.

Because the acceleration of substitution


cont. ls wouild reduce the stock of ODS- aggressive phaseout but exempted servicingusing equipment requiring servicing after use for refrigeration and air conditioningthe phaseout date, the need for retrofits and equipment. Enough ODS use was allowed forearly retirements was reduced. This servicing after the phaseout for newstrategy, therefore, lowered the present equipment to prevent the expensive retrofitsvalue of total phaseout costs as compared and early retirements that would otherwise beto the allowable strategy. required.

Reference. ICF Incorporated modeled thisstrategy to follow closely the phaseoutactions recommended to the Thai The unit abatement cost of each alternativegovernment, most of which were later technology modeled, that is, the cost peradopted in the National Phaseout Policy. kilogram of ODS reduction achieved, includedLike the national policy, this strategy incremental capital costs, variable costs, andassumed that implementation of some nonrecurring costs. Costs were incrementalalternative technologies would occur more because they were defined relative to therapidly than the allowed or forward-looking baseline (that is, incremental to the cost ofstrategies because of global trade and continuing to use ODSs).3 As a result, theenvironmental pressures. In part, this aggregate costs estimated in the analysis weredevice allowed us to circumvent the issue the net social costs of the phaseout.of variable chemical prices in the unit Because Thailand has no ODS producers,abatement costs. A user defined strategy incremental phaseout costs can be incurredwas also deemed necessary because several only by (1) manufacturers who produce ODS-criteria could be important to a risk averse containing equipment or who use ODS in thedecisionmaker forming national phaseout manufacturing process and (2) consumers ofpolicy, yet could be difficult to quantify in products made with or using ODS. Hence,a modeling framework. These include: costs in the model were divided into industry

* human health and environmental benefits costs and consumer costs.resulting from an earlier phaseout; Because industry can be expected to pass

some of its increased costs to consumers in the* competitive advantages from manu- form of higher prices, we sought to avoid

facturing facilities that use state-of-the-art double counting costs in the simulation. Coststechnologies; were evaluated at the point where they were

* the value to producers of exports to markets first incurred. We believed this to bewhere consumers prefer products appropriate because various factors couldcontaining ODS substitutes (especially if limit the scope of price increases byODS-containing products must be labeled); manufacturers in the near term. Where it was

* international goodwill fostered by anticipated that some pass-through wouldThailand's taking an environmentally pro- occur, the burden of incremental costs wasactive position; and split between manufacturers and consumers to

avoid double counting.* the potential for earlier phaseout dates Examples of specific industry costs

mandated by the parties to the Protocol. modeled were:

The reference strategy was based on theresults of the other strategies, and representeda compromise between the accelerated 3. Depending on the alternative and the sector, costsstrategy and the allowable or forward-looking can either be positive or negative, where negative costsstrategies. The reference strategy called for an indicate economic benefits of phasing out ODS use.


Capital costs, primarily the incremental coscoMby "levelizing" the cost components.costs of new or modified facilities and Levelizing the costs-that is, creating aequipment. An example in the model was smooth annualized stream from the presentthe cost of new technologies, such as an value of the costs over time-was necessary toaqueous cleaning machine assumed to avoid double discounting when total phaseoutreplace a CFC- 113 vapor degreaser or a new costs were calculated.off-site recycling plant. Calculation of unit abatement costs also

Variable costs or operating costs, such as accounted for the fact that not all end uses

raw materials, labor, and energy expenses. would have the same profile of chemical use.Examples of specific variable costs Some end uses modeled would be expected tomodeled were increased chemical costs, have only a one-time demand for an ODS (forincreased labor charges associated with example, HFC- 1 34a used to chargeimproved charging and maintenance refrigerators manufactured for export), somepractices, and incr-eased electricity and end uses would have an annual demand

. ~~(CFC-1 13 use In an electronics cleaningwater consumption associated with the use (F- 3uei neetoiscennofwaltercona ptiv facility), and some end uses would have

uneven demand over time, which implies it* Nonrecurring costs or fixed costs incurred was either not constant on an annual basis or

on a one-time basis. Examples used in the that would occur only periodically. Examplesmodel included personnel retraining costs of the latter include refrigerators and MACs

and research and development for adapting that in the simulation were "charged" withnew technology. HFC- 1 34a initially and then serviced every few

Examples of costs borne exclusively by years, depending on their respective servicingconsumers included: frequencies.

tUnit abatement costs were calculated* Incremental operating costs such as

increased electricity costs associated with assuming a discount rate of 12 percent.operating an HFC- 1 34a refrigerator. Estimating Premature Retirement and

* Forced retrofits or early retirement costs Consumer Costssuch as those arising from shortages in thesupply of ODSs. These would forceconsumers to bear the costs of retrofitting constraints imposed by the Montreal Protocol

or retiring their mobile alr conditioners and do not allow sufficient quantities of ODS to beor refiringeatoir earler than conditionecessa imported to meet the total servicing demandrefrigerators earlier than necessary. fo rergrto an arcnioigfor refrigeration and air conditioningThe costs associated with each control in equipment in a given year, the excess demand

the model were not consistent over time. must be eliminated by retiring or retrofitting aSome nonrecurring costs, such as capital certain fraction of the equipment stock. In thecosts, R&D, and training, were expected to be model, enough retrofits and retirements wereincurred once for all units on a plant basis, implemented when necessary so that ODS

while other nonrecurring costs would typically demand equaled ODS supply. Retrofit andoccur only once on a unit basis, such as the retirement controls reduce future demand ascost of a compressor for an HFC-134a well because they reduce the stock ofrefrigerator. Incremental operating costs equipment that will need service in the future.(energy, labor, chemical costs, and so on) The incremental costs associated with retrofitswould be incurred over time. These disparate and premature retirement were calculatedcost profiles for individual controls were made based on the actual stock predicted to becomparable on a per-kilogram basis in retired or retrofitted.

70 OZONE LAYER PROTECTON: COUNTRY INCREMENTAL COSTS

The calculated reductions achievable Technical Options and Economicthrough retrofits and premature retirements Assumptionsand the costs associated with these activitiesdepended on projections about the size of the Tcnalptions a nd economicequipment stock and the number of units assumptions about the alternative technologiesrequiring sevce. aOSCtMeestimated.thes used in the model are summarized in Tablefactorsqusingth servic ingco m feq ten these 4-4. These were obtained from a variety offactors useng the servicingo not soc by sources, including trade literature, govenmentend use and the distribution of the stock by aece,itroenetlognztos

aeor vintage. The vintaging information was agencies, intergovernmental organizations,agelo vintage. the stock istrmation was chemical suppliers, ODS users, tradealso used to adjust the stock distribution when associations, independent consultants, andpremature retirements or retrofits were academics. The data collection was asundertaken. The servicing frequency wasmade a variable that could be adjusted for exhaustive as possible withi the existing costsensitivity analyses. constraints.

Retrofits and early retirements wereassumed to occur only when equipment Results and Recommendationsrequired service. As a result, retrofits andretirements in the model that occurred after this sei summa ri terss ofcthestudy as well as ICF Incorporated's recoin-the phaseout date were made to take place mendations for the phaseout of ODss inover a period of time based on the servicing Thailand.frequency assumed in the end use. Somestudies we have seen assume that all Summary Incremental Costs-1991 toremaining equipment will be retired or 2010retrofitted as soon as ODSs are no longeravailable (that is, in 2010). This can result in As noted above, incremental costs are thea substantial overestimation of costs because costs above those that would be incurred in(1) equipment without further need of service the absence of the Montreal Protocolcan be normally retired and (2) retrofit and provisions. Costs for each strategy discussedretirement costs incurred after the phaseout below are the present value of incrementalyear will be discounted more than those costs predicted to be incurred between 1991incurred in the phaseout year. More and 2010. Estimates do not include the costsrealistically, equipment should be retired or associated with promotion of the ODSretrofitted in a simulation (and in the real phaseout, training, and institutional changes.world) only when it requires servicing. The Costs for the four strategies modeled areonly way in which all CFC-using equipment summarized in Figure 4-1 and describedwould be retired during a single year is if it all below.required annual service. This is not the casefor refrigerators, which are serviced Accelerated Strategyinfrequently, and which are usually the most The incremental cost of complying with theexpensive item to retire prematurely per Montreal Protocol under the acceleratedkilogram of ODS use avoided. Retiring all strategy was estimated to be $471 million.refrigerators as soon as ODSs are no longer Over 94 percent of the costs were attributableavailable results in a substantial over- to phasing out domestic ODS use. About 90estimation of the costs of premature percent of the costs would occur in the airretirement. The same is true in the case of conditioning and refrigeration sectors,retrofitting MiAcs. primarily because of their high per-kilogram

control costs. Because conversion to


alternatives was predicted to occur early in under the allowable strategy. Although thethis strategy, substitution costs were incurred strategy was similar to the allowable strategy,over the entire range of the analysis. Costs the optimization of refrigeration and airfrom retrofits and premature retirements, conditioning controls limited retrofits andhowever, were minimal. Retrofit costs early retirements to 17 percent of domesticaccounted for only 6 percent of all domestic phaseout costs. The forward-looking strategycosts. The accelerated strategy would result in would result in about 550,000 metric tons ofabout 618,000 tons of avoided ODS use at an ODS use avoided at an average cost of $512average cost of $763 per ton. per ton. This was the least expensive strategy

modeled for Thailand, both in terms of theAllowable Strategy total costs as well as the cost per kilogram.

The incremental cost of complying with theMontreal Protocol under the allowable Reference Strategystrategy was estimated to be $431 million, $40 The incremental cost of complying with themillion less than under the accelerated Montreal Protocol under the reference strategystrategy. Many of the costs would be incurred was estimated to be $418 million. About 94far into the future, which reduced their present percent of the cost was attributable to phasingvalue when they were discounted. About 96 out domestic ODS use. Particularly importantpercent of the cost was attributable to phasing were the refrigeration and air conditioningout domestic ODS use. Air conditioning sectors, which had high per-kilogram controlaccounted for 49 percent of total domestic costs and account for a significant percentagecosts and refrigeration for another 47 percent. of total ODS consumption in Thailand. UnderThe costs for these sectors was high primarily this strategy, phaseout of all ODS use wouldbecause of retrofits and premature retirements occur by January 1, 1997, with the exceptionpredicted to occur during the period of CFC use for servicing air conditioning and

2005-2015. Although the total cost of the refrigeration equipment existing at that time.allowable strategy was similar to the total cost CFC use for servicing equipment wouldof the accelerated strategy, a much higher continue until 2010 if the reference strategypercentage of the cost came from retrofits and were followed. The reference strategy wouldearly retirements. In the allowable strategy, 76 result in about 605,000 metric tons of avoidedpercent of the costs came from retrofits and ODS use at an average cost of $690 per ton.early retirements arising from the lateimplementation of substitution controls. Sensitivity Analyses

Although the magnitude was simlilar, the The estimated cost of the ODS phaseout wasdistinction is important for equity reasonsbecause retrofit and retirement costs would snsitv tooaenumber of u assuptOnSprimarily fall on consumers. The allowable including projections of unconstrained ODSstrategy would result in about 542,000 tons of demand, technical assumptions underlying thestravoidegy wOulD ret in aboutrage000 tons of $costs of ODS control options, chemical prices,avoided ODS use at an average cost of $796 discount rate, and constraints imposed by theper ton. Montreal Protocol. We changed the phaseout

Forward-Looking Strategy scenario by varying some of these in differentmodel runs.

The incremental cost of complying with theMontreal Protocol under the forward-looking Unconstrained ODS Demand Projectionsstrategy was estimated to be $282 million.This was $189 million less than under the Prections of uncnr ODS demand

accelerated strategy and $149 million less than


economic growth rate assumptions for as soon as possible, even with constantThailand. Given the narrow separation chemical prices. Second, incremental costs, asbetween the two projections provided by the we have construed them, are based on base-Bank of Thailand, we concluded that the cost year chemical prices, even if the controls areestimates were not particularly sensitive to chosen using price projections.projections of unconstrained demand and onlythe low growth projections were used. Discount Rate

The choice of a discount rate has anTechnical Assumptions Underlying ODS Tecoc fadson aehsaControl Options important impact on incremental costs

because the present value of costs for anyNo sensitivity analysis was performed for specified strategy will increase as the discount

the technical assumptions underlying ODS rate is lowered. All things being equal (ceteriscontrol options because the assumptions used parabus), incremental costs can bewere based on actual data available from substantially higher if a lower discount rate isThailand or documented from investment used. The 12 percent real discount rate used inprojects undertaken in other developing the base case was essentially the private ratecountries. of return in Thailand at the time of the study.

ICF Incorporated examined the results using anChemical Price Projections alternative discount rate based on the public

Changes in the prices of ODSs and substitute rate of return. In June 1991 the medium termchemicals attributable to the global phaseout rate for goverment-issued bonds in Thailand

was 11 percent, and the inflation rate was 6will have substantial impacts on the costsfaced by industry and consumers. As noted percent, yielding a 5 percent real rate of returnabove, we originally modeled the Thai for public investments. This was the discount

situation using constant real chemical and rate used for purposes of sensitivity analysis.situation using constant real chemical and Asmayo eut ae nti aeisubstitute prices. A more sophisticated A semmary of results based on this rate 4sapproach would have been to develop the presented in Table 4-2.reference phaseout strategy using unit As expected, the reference and accelerated

strategies showed an increased cost but noabatement costs based on chemical pricegchange in the total amount of ODS used vis-a-

strategy would be calculated using unit vis the results calculated using a 12 percentabatement costs reflecting base year ODS discount rate. The same was true for theprices. This methodology allows for the allowable strategy. This outcome arose

because the lower discount rate did not changedevelopment of a realistic phaseout scenario teodrno h otosi h aeowithout omitting the imposed costs of the the ordering of the controls In the case ofProtocol. Thailand, and therefore, the same sequence of

Although using a price projection can controls was predicted as when using a 12change the timing of controls to a certain percent discount rate. The forward-lookingextent, the impact on total incremental costs is strategy, however, showed an increase ofgenerally modest for two reasons. First, about 5,000 tons of ODS avoided, indicating asignificant changes in costs are usually change in the order in which controls wereassociated with changes in the number of implemented. As shown in Table 4-2, aretrofits and premature retirements. However, substantial decrease in the percentage share ofrising chemical prices are not likely to change predicted premature retirement and retrofitthese numbers significantly, since substitution costs also occurred with a lower discount rate;controls in serviced end uses generally occur this indicates that these costs, although they


would be incurred 15 to 20 years in the future, the costs for different strategies, assumingbecame more important. different ultimate phaseout dates.

Figure 4-I shows the costs associated withConstraints Imposed by the Montreal the three strategies for phaseout dates rangingProtocol from 1997 to 2010. The costs for a 2010

Because the Copenhagen meeting of the phaseout were the same as those presentedparties was imminent at the time the study was earlier. The top panel in the figure shows theconcludeds ICF Incorporated analyzed the sensitivity of costs to the phaseout date,concluded, of corpor canalyzed the assuming a 12 percent discount rate, and the

sensiivityof csts t chages i the bottom panel shows the sensitivity, assumingconstraints imposed by the Montreal Protocol. . ' sWith no prior knowledge of changes in a 5 percent discount rate. A lower discountinterim targets, we examined the sensitivity of rate not only results in substantially higher

total incremental coststodifferentphaseoucosts overall, but also results in the allowabletotal incremental costs to different phaseout staeyawy en oeepnietadates, keeping all applicable interim targets evente aleated phae staegy.the same as for the 2010 phaseout. The costs IDfor each approach were recalculated assumingdifferent phaseout dates, that is, dates by Recommended Plan of Actionwhich all import and use of ODS, including for ThailandODS required to service existing airconditioning and refrigeration equipment, was The study team investigated a number ofno longer possible. ways that Thailand could meet its phaseout

This analysis was conducted in case the obligations, and recommended that thephaseout dates were revised in Copenhagen, in country commit to a rapid ODS phaseout aswhich case the new dates would be binding on part of its national industrial policy. OurThailand. The same constraint would apply if suggestion was that the government move toThailand unilaterally adopted a schedule that eliminate the use of ODSs in new equipment bycalled for a faster phaseout. 1997 while allowing consumption of small

If the date of the ODS phaseout were earlier quantities of CFCs for servicing certain long-than 2010, the costs incurred under the lived refrigeration and air conditioningforward-looking, allowable, and accelerated equipment up until the final phaseout datestrategies would increase. Even though an allowed by the Protocol.accelerated strategy would achieve a phaseout As discussed above, our rationale forof all ODSs by 1996, except for ODSs required recommending this overall strategy was that ato service existing equipment, a phaseout date rapid phaseout would benefit Thailand inof 1997 that did not permit use of ODSs even ways that were not captured by the economicfor servicing uses after that date would result modeling techniques used. Implicit in ourin substantially higher costs because recommendations were the probable costretrofitting or prematurely retiring existing increases in ODSs and cost decreases ofequipment would start in that year. Premature alternatives. Furthermore, human health andretirement and retrofit considerations would environmental benefits, competitivebecome important sooner; and the conversion advantages, protection of markets, andto alternatives, instead of being pushed off international goodwill are real benefits, ifuntil later in the decade, would have to begin difficult to quantify, that we believe should beearlier. Because the phaseout date, by which factored into the analysis.no use of ODS is allowed (not even for Our recommendations were grouped intoservicing existing equipment), was made a four categories: institutional and policyvariable in the model, it was easy to compare measures, regulatory measures, financial


measures, and sectoral measures. Table 4-3 information on ODS alternatives, costs ofsummarizes these measures. conversion, and so on. As a regulatory

measure, the research team recommended thatInstitutional and Policy Measures the government strongly consider product or

We recommended that the Cabinet should process bans for those uses where substitutesor ODS are clearly available, as it had done

commtt Thailand to a rapld phaseout, rltie with moderate success with the ODSs in theto that allowed a developing country by the arsl etrProtocol. This recommendation reflected our aerosols sector.

concern that earlier phaseout actions inThailand suffered from the sense that the issuewas not a government priority. The Study ODS substitutes are generally still moreTeam also recommended that the Office of the expensive than the chemicals they replace;Board of Investment make formal its policy of this was certainly true at the time of the study.not extending Board of' Investment (BOI) To make substitutes more attractive and topromotional privileges to ODS-consuming encourage recycling of ODSs, ICF Incorporatedindustries, and that within the Department of recommended that the government increaseIndustrial Works, a special unit be set up to the duty on ODSs and waive or reduce dutiesdeal exclusively with the ODS phaseout, on chemical alternatives, possibly in aprimarily determining government policies, measured way to prevent a sudden increase inmonitoring consumption, and making sure that prices.Thailand meets its commitments as a party tothe Protocol. Other recommendations were Sectoral Measuresthat the Industrial Finance Corporation of Refrigeration and air conditioningThailand, with its experience in providingfinancial and technical assistance to Thai Delaying the phaseout in this sector couldindustries, be given authority to manage and impose very large costs on Thailand, asdisburse funds received from multilateral and equipment would need to be retrofitted orbilateral sources, and that the Department of retired prematurely after the phaseout date.Industrial Works and the Customs Department Our main recommendation in this sector wasjointly undertake a review of ODS data that the government require manufacturers ofcollection procedures. household refrigerators and MAC systems to

convert to the refrigerant HFC- I 34a as soon asRegulatory Measures possible. ICF Incorporated also recommended

The Study Team suggested that the that the government expand its existingDepartment ofIdusra orkrecycling program for MACs and institute oneDepartment of Industrial Works set upy g g

for commercial refrigeration and airsubsector working groups consisting of conditioning systems, as well as ban therepresentatives from industry, government iagencies, and academic institutions that would import of building chillers that could operateadvise the government on its ODS phaseout only on CFC refrigerants.policies. We believed that Thai industry, in Solvent cleaningparticular the Federation of Thai Industries,could play a major role in the switch to ICF Incorporated's recommendations in thisalternatives. Within the Industrial Finance sector were that Thai industry install solventCorporation of Thailand, we recommended recycling equipment and adopt housekeepingthat a resource center or clearinghouse be set measures that reduce solvent consumption inup with access to databases containing the short term. We also suggested that the

government persuade multinational


corporations to eliminate ODS solvent use in Drawing LessonsThailand at the same rate that they do so in from the MethodologyJapan, the United States, and Western Europe.The research team also suggested that the Lessons about Methodology to Benefitgovernment discourage or ban the use of Future Studiescertain HCFCs as replacements for CFC- 113 in

solvent cleanig applicationsAs mentioned above, one way to improvesolvent cleaning applications, the methodology Lised in this study would be

Foam blowing to base the model decisions on priceprojections for ODSs and ODS alternatives. This

In the foams sector, ICF Incorporated would allow the model to approximate ansuggested that Thailand's Department of optimal strategy more closely. AnotherIndustrial Works ban the use of ODSs as foam improvement might be to add the option ofblowing agents in certain applications and incorporating projected price increases in thework with industry to make the conversion to baseline. If run with base year prices, thealternatives, such as methylene chloride and model would produce estimates of the totalpentane. We suggested that the Department social costs of the Protocol. If run withclosely monitor developments regarding projected price increases, the model wouldalternatives for CFC-11 blowing agents in produce estimates of the voluntary costs ofinsulating foams, and ban the use of CFC- 1 complying with the Protocol. This is the samewhen an adequate alternative is found. as usilg a post-Protocol baseline, and the

voluntary costs are the same as the total socialAerosols costs minus the imposed costs of the Protocol.

Although most use of ODSs in aerosols is In our experience it has been difficult toAlthough most use of fndODSDpiceSrojctins aerosolsch btisprohibited, anecdotal evidence pointed to tfind ODS price projections that are much bettercontinued use in some nonmedical than guesses. One possible solution wouldbeapplications. ICF Incorporated recommended to have the model generate its own pricethat the Department of Industrial Works work projection based on the unit abatement costswith manufacturers and chemical suppliers to of the controls required to meet the interimwithmanfacures ad cemial uppier to targets each vear. However, there are adetermine the exact consumption in this sector g yand take measures to ensure a speedy number of difficulties in this approach. First,conversion to alternative formulations. ODS prices cannot be considered in a closedFire Protection system. Prices will be influenced by global

market forces and by unforeseeableHalon use in Thailand is not large, but these technological factors such as the development

chemicals have high ozone depletion of a drop in substitutes or inexpensivepotentials. ICF Incorporated recommended that retrofits. Second, prices projected in thisthe government ban the manufacture and sale manner are sensitive to assumptions about theof hand-held portable halon 1211 fire cost and availability of controls. Third,extinguishers, require recycling of halon 1211, projecting prices would require complicatingand ban tests of halon 1301 total flooding the analysis with modeling of ODS stockpiling.systems. Finally, this approach does not take account of

elasticities in ODS demand, which may limitthe range of possible price increases.

The strategies defined by the World Bankare appropriate and should continue to beevaluated. Although different studies presentthe descriptions of these scenarios in different


forms, in essence, three scenarios are usually collected from different sources was oftendescribed. These are the accelerated strategy, inconsistent or contradictory.which is intended to show the fastest possible The best, most consistent information wasphaseout without regard to any institutional or obtained from ODS importers and distributors,capital constraints; the allowable strategy that who identified the consumption of each ODSshows the drawbacks of doing the least by end use in the Thai economy. Data waspossible in any year while meeting the harder to obtain from enterprises that usedProtocol's requirements; and the optimized ODSs, particularly smaller users. In areasstrategy that presents, as the name suggests, where the Thai government had alreadythe minimum costs that a country will likely adopted regulatory measures limiting ODS use,have to incur. Although the optimized strategy such as in aerosol applications, sound datacaptures most of the problems associated with was extremely hard to obtain.delaying actions to eliniinate ODS use-chief Regarding unit abatement costs,among them is the avoidance of premature nonappliance insulating and noninsulatingretirements and retrofits-there are certain foams and halons presented a challenge, asconsiderations not captured by such a strategy, there was less information available onsuch as environmental and trade benefits. In conversion costs. In the Thailand study,addition, an earlier phaseout for use in new fortunately, these end uses were not critical.equipment could be mandated by the parties, Additional data collection is also justified inwith use for servicing permitted until the the area of determining servicing requirementsoriginal 2010 phaseout date. In either case, and lifetimes of refrigeration and airthere exists the need to add a preferred conditioning equipment in developingstrategy, much like the reference strategy countries, where age profiles, servicingdeveloped by ICF Incorporated for Thailand practices, use patterns, and operatingand other countries like Poland, the conditions often differ greatly from those inPhilippines, and Taiwan. the developed countries. This is particularly

important as the costs of premature retirementAreas of Policy Research and Data are a large component of incremental phaseoutCollection costs for most strategies.

As noted above, ODS and alternative price Usefulness of Country Incremental Costtrends figure prominently in determining a Methodologycountry's optimal phaseout strategy. We werehampered somewhat in the study by the The main strength of the methodology usedunavailability of reliable price projections. for analyzing the country incremental costs isAdditional research on global supply and that it lays out a sequence of actions thatdemand and projected prices for ODS and illustrates some general principles useful inchemical alternatives, particularly HCFCs and designing policies to achieve an ODS phaseout.HFCs, is warranted by the important effect they The model illustrates, for example, thewill have on phaseout paths and costs. importance of avoiding retrofits and premature

Data collection was a problem in some retirements through early conversion toareas of the study. Identifying sources of data alternatives in new refrigeration and airwas not difficult, but it proved particularly conditioning equipment and throughtroublesome to persuade companies to part allowances for a servicing tail. The model alsowith closely held information, particularly as illustrates the importance of comparingthe study was being conducted for a control measures on the basis of unitgovernment ministry. As expected, data abatement costs throughout the entire life of


equipment rather than at just one point in Making Calculations of Incrementaltime. Costs More Transparent

In addition, the methodology is useful The model developed by ICF Incorporatedbecause it calculates social costs of complying is a spreadsheet framework consisting ofwith the Montreal Protocol, which may serve several linked worksheets. Although thisas a lower bound estimate of reimbursement seralmdwokht.Atouhhsnecessalery froud thestimateofreimburd hent approach makes it difficult to model variousnecessary from the Multilateral Furd. The features such as vintaging, prematureincremental cost estimate is a lower bound retirements and retrofits, and implementationbecause the model does not account for costs of recycling and recovery controls (includingassociated with conducting the various the interaction with other controls), the useranalyses, feasibility studies, and institutional benefits from the transparency that comesstrengthening activities that will be required.It should be noted that the methodology is not oith abate form individualcuont

useful fr determning thelevel of of unit abatement costs for individual controls,useful for determining the level of reutin in. Sahee nulyreimbursement appropriate for specific premature retirements and retrofitsprojects. It should also be noted that decisions unertake sequens andabout strategies to pursue should be based not investmen, anduincrntal costs as donly on total incremental costs, but also on i th,pang uncrem al transparenquantity of ODS use avoided; average cost per X th model us by Ss IreoratedaEen itunit of ODS eliminated; impact on consumers; i

. .' . . ' ~~~macro algorithms used to calculate thedifferential impact on sectors; logistical and mcoagrtm sdt aclt hdifferentioal ipctonsidecatiors; log uica a different strategies are available for the user toinstitutional considerations; and equity review and customize.concerns.

-100

Table 4-1: Summary of Model Results for Various Phaseout Strategies

Premature retirement

Phaseout date" Average cost Domestic costs Export sector costs and retrofit costs Industr costs Consuimer costs

Total ODS use of ODs use % of % of

Total costs avoided" avoided total % of total domestic % oftotal % of total

Strategy Annex A ODS Annex B ODS (mm $) (tons) ($/ton) mm $ costs mm $ costs mm $ costs mnm $ costs mm $ costs

Reference 1997 1997 418 605,000 690 393 94% 25 6% 22 5% 283 68% 134 32%

Forward 2 0 10c 2005 282 550,000 512 265 94% 17 6% 72 27% 181 64% 101 36% N

looking0

Allowable 2010 2005 431 542,000 796 415 96% 17 4% 314 76% 167 39% 264 61%

Accelerated 1996 1996 471 618,000 763 443 94% 28 6% 18 4% 314 67% 158 33%

a Phaseout date indicated is for all ODS use except for use in servicing existing equipment. All ODS use ceases in 2010.

b Total ODs avoided is shown to the nearest thousand metric tons.

c cFc use, except for servicing existing equipment, ceases in 2008.

Note: (I) All costs are present values discounted at 12 percent. 0

(2) Numbers may not add due to rounding. C

1-

,0n

Table 4-2: Summary of Model Resultsfor Various Phaseout Strategies EU

(Using a 5% Discount Rate)

Premature retirementPhaseout datea Average cost Domestic costs Export sector costs and retrofit costs Industry costs Consumer costs

Total ODS of ODS use % of % of % of % of

Annex A Annex B Total costs use avoidedb avoided total total domestic total % of total 0Strategy ODS ODS (mm $) (tons) (S/ton) mm $ costs mm $ costs mm $ costs mm $ costs mm $ costs 9

Reference 1997 1997 720 605,000 1,189 680 94% 40 6% 51 7% 466 65% 254 35%

looinagd 2 0 10 C 2005 539 555,000 971 509 94% 31 6% 84 17% 323 61% 211 39%

Allowable 2010 2005 974 542,000 1,796 943 97% 31 3% 718 76% 393 40% 581 60%

Accelerated 1996 1996 777 618,000 1,257 734 94% 43 6% 43 6% 493 64% 283 36%

a Phaseout date indicated is for all ODS use except for use in servicing existing equipment. All ODS use ceases in 2010.

b Total ODS avoided is shown to the nearest thousand metric tons.

C CFC use, except for servicing existing equipment, ceases in 2008.

Note: (I) AD costs are present values discounted at 5 percent.(2) Numbers mnay not add due to rounding.


Table 4-3: Recommended Action Plan

Year Recommendation

1992 Make a Cabinet-level commitment to a rapid ODS phaseout part of Thailand's industrial policy.

Stop extending Board of Investment promotional privileges to new ODS-consuming industries.

Set up a special unit in Department of Industrial Works (DIW) devoted to ODS phaseout activities.

Give IFCT responsibility for managing and disbursing funds received for phaseout related projects.

Impose a progressively higher excise tax on imported ODS.

1993 Undertake a public awareness/industry outreach campaign about the ODS phaseout.

Set up subsector working groups to advise the government on ODS policy.

Set up in IFCT an ODS resource center with access to the latest information on alternatives.

Improve DIW and Customs Department data collection methods and computerize import records.

Ban tests of halon 1301 total flooding systems in which the halon is released.

Remove duties on ODS recycling equipment, alternative technologies, and chemical substitutes.

Undertake a pilot project promoting recycling of refrigerants lrom commercial refrigeration systems.

Ban the import of chillers that can only use CFC-l I or CFC-12.

Continue testing the HCFC-22/HCFC- I24/HFC- I 52a ternary blend in mobile air conditioning systems.

Encourage industry to adopt measures that reduce solvent consumption.

Adopt measures to limit the use of HCFC-123 and HCFC-141b as solvents.

Use the DIw/IMTI/EPA program to compel Thai companies to eliminate solvent use.

Institute a CFC- 1 2 recycling program for mobile air conditioners.

Require remaining aerosol manufacturers to adopt non-ODS technology.

1994 Ban the manufacture, sale, and import of hand-held portable ha]on 1211 fire extinguishers.

Require recycling of halon 1211 when units are serviced.

Require household refrigerator manufacturers to convert to HFC-134a systems.

Require automobile assemblers to convert to mobile air conditioners that use HFC- I 34a.

If necessary, implement a quota system that will keep imports below targets.

Ban the use of CFC- 12 in polystyrene foam production.

1995 Ban the use of ODSs as a foam blowing agent in packaging foam and flexible polyurethane foam.

Ban the use of CFC- I I in flexible polyurethane foams.

1997 Ban the use of CFC- 1 I in rigid insulating foams.

Table 4-4: Characteristics of Control Technologies Use in the Analysis

Earliest etimnated Years to reacht ma. tunurn Maxinmun mark-et

Secror Application Contrrol technology aiailabditics penietrationi penetration()

Refngerafion Homie appliances HiFc-1I34a 1993 3 100

Conmiercial HCFC-22 1993 3 30HIFC-1I34a 1993 3 70Recovery 1993 3 50

Air condidoning Mobile air conditioning i-4FC-1I34a 1993 3 100

Recycling 1993 4 50

Retrofits (HFC- I34a) 1994 1 100

Building chillers HCFC- 123 1993 2 90

HFc-134a 1993 2 10Recovery 1993 3 100Retrofits (HCFC-123) 1993 1 90

RetrOfitS (H-FC-1I34a) 1993 I 10

CJFC-1 13 solvent cleaning Electronics/precision cleaning No-clean 1993 4 40 1Aqueous 1993 4 20 S

Serr.-aqueous 1993 4 30 c

HcFC-225 1995 2 10 :

Recovery 1993 2 100

Metal cleaning Aqueous 1993 4 40Semii-aqueous 1993 4 60Recovery 1993 2 100

l1,1l-tmchloToethaTne Electronicstprecision cleaning No-clean 1993 4 403

solvent cleaning Aqueous 1993 4 200

Semni-aqlueous 1993 4 30HCFC-225 1995 2 10 -

Recovery 1993 2 100E

Metal cleaning Aqueous 1993 4 40Semi-aqueous 1993 4 60Recovery 1993 2 100

0

Aerosols Cosmetics/cleaners Hydrocarbons 1993 I 100

0

Foams Appliances HCFC blend 1994 220Other 1994 2 so

Flexible Methylene chloride 1993 2 100

Rigid HcFC blend 1994 2 100

Polystyrene packaging H4CFC-22 1993 2 30other 1993 2 70

Fire protection Portable Chensical replacenients 1995 4 20

(halon 121a1) Altenatives 1993 2 80

Total flooding Chenical replacensenis 1995 4 20

(halon 1301) Alternatives 1993 2 80

Alrcondiuoning Mobdeaircnndiuoning HFC-134a 1493 A 100 n~~~~~~~~~~~~~~~G


Figure 4-la: Phaseout Costs vs. Phaseout Year

1 6 00

Forward Looking

1,400 ~ - - - - - - - -~ - _ _ _ _

Ailowable

c' 1,200 - -_ -- Accelerated

CS)

0 80001

a 600-X

Cd)

400- - --

200 - -1997 1 999 2001 2003 2005 2007 2009

ODS Phaseout Year

Figure 4-11b: Phaseout Costs vs. Phaseout Year

2,000T,_ _

1 800t _ Forward Looking

t 1 600 __-_X_ Allowable

) 400 Accelerated

1,200 t -- 21000-- ----------- __ __ X-

0 800----- -

o 600 -- ---- - --

(L 4004 -

200 >---- ----

1997 1999 2001 2003 2005 2007 2009ODS Phaseout Year

5

Evaluation of Incremental Cost Methodologiesfor Phasing Out Ozone-Depleting Substances:Case Studies of Tunisia and Czechoslovakia

Ulla Blatt Bendtsen

Background

THE WORLD BANK and the Global Environment Facility areevaluating the economic and policy issues of Country Programsfor implementing the Montreal Protocol in developing countriesand countries in economic transition, such as Central andEastern Europe (CEE) and the Commonwealth of IndependentStates (cis). Incremental cost estimates and policy recommenda-tions are features of the ozone-depleting substances (ODSS)Country Programs that form the basis for financial transfersfrom the Multilateral Fund for the Implementation of theMontreal Protocol and the Global Environment Facility (in thecase of economies in transition).

83


Although this evaluation specifically * Step 3: technical assessment of the mostaddresses the problem of calculating the appropriate substitution techniques for eachincremental costs of phasing out the use of ODS application.ODSs, it is part of a broader effort by the * Step 4: estimation of the incremental userWorld Bank to address economic and policy costs of each replacement technique.issues arising in global and regionalenvironmental contexts. * Step 5: estimation of the incremental

This chapter will form part of an integrated consumer costs of premature obsolescencereport on costs and policies related to of refrigeration equipment.elimination of ODSs. The report comprises * Step 6: specification of the constraintsanalytical papers on theoretical and conceptual imposed by the Montreal Protocol as wellissues related to international funding of ODS as potential technical and institutionalphaseout projects and methods for estimating constraints.country incremental costs as well as country * Step 7: specification of alternative ODScase studies which focus on methodologiesused in practice to analyze cost-effectivephaseout strategies. a Step 8: calculation (minimization) of the

This chapter comprises the country case incremental costs of each phaseout scenariostudy on Tunisia and Czechoslovakia. The relative to the without Protocol situation.objective of the chapter is to evaluate the * Step 9: cash flow analysis.method used for calculating incremental costsof ODS phaseout in these countries and to * Step 10: sensitivity analyses.discuss the recommended national policies. Steps I through 7 comprise the required

This chapter describes the methodology input data to the chlorofluorocarbon (cFc) costused by cowiconsult to analyze cost-effective model which was developed by cowiconsultODS phaseout strategies in a number of with the purpose of estimating a country'scountries, including developing countries, minimum incremental costs by phasing out thecountries in economic transition, and the use of ODS subject to the requirements of theEuropean Community. This chapter focuses Montreal Protocol or any alternative phaseouton the methodological and conceptual issues, schedule. The model, a linear programmingillustrated primarily by the experiences of the model, is described below under the sectionTunisian Country Program preparation. Case entitled Model for Calculating Incrementalstudy material from Czechoslovakia is only Costs of ODS Phaseout. Steps 8 through 1O arereferred to when this illustrates a special issue the model calculations. A brief review of eachor problem, for instance in relation to the early step is provided below.phaseout date (1996) required in CEE and CIScountries. The chapter is based on work Steps in Estimatingalready completed (see Table 5-1 for a list of Country Incremental CostsODS references). No new data gathering,research, or analysis has been made. Step 1: Assessment of present and past ODS

The steps used to estimate the country consumption by specific ODS applicationincremental costs of ODS phaseout in Tunisiaand Czechoslovakia are summarized below. The consumption of ODS is divided into six* Step 1: assessment of present and past ODS major user sectors (in the model denoted: H.,

consumption by specific applications. for n = 1,...,6):

* Step 2: projection of unconstrained ODSdemand, 1991-2010.

5. Evaluation of Incremental Cost Methodologies for ODS Phaseout Case Studies 85

* HI: refrigeration, air conditioning and use of ODSs for basic domestic needs. Thus,heat pumps; the use of ODSs in export products must be

* H2 : rigid foam; regulated according to the same phaseout* H3: flexible foam; schedule as in developed countries.* H4: aerosols, sterilants, and other uses; Estimates of ODS consumption were made

* H 5: solvents; and for the period 1986-1991 based on import

* H6: fire extinguishants. statistics, previously conducted surveys, and

interviews conducted during field missionFor each major application area, a set of with the major ODS importers, distributors,

subapplications is defined, depending on the and users.

specific consumption pattern of the country.

These are denoted HnSi for n = 1,...,6 and Step 2: Projection of unconstrained demand,i = 1,.. .,6 and comprise, in the case of Tunisia 1991-2010

H Si (i = 1,.. ,6): The projection of unconstrained demandHlS (i ' ',.. .,6): assumes that the Montreal Protocol did not

* domestic refrigeration for the home market; exist, that is, that ODS are not regulated (in

Tunisia or elsewhere), and unconstrained- domestic refrigeration for the export quantities of controlled substances continue to

market; be available from existing sources at current

* servicing of domestic refrigerators; prices. The base for projecting unconstraineddemand is the 1991 consumption adjusted for

' commerialrfrigertionpossible reductions in ODS use that have* industrial refrigeration; and already taken place. In Tunisia, for example,

.mobile air conditioning. part of the aerosol sector had already shifted toalternative propellants, such as liquid

H2S (i = 1,...,4): petroleum gas (LPG). In the model, DO(H,S)denotes the demand for ODS in year 0 ( 1991)

i polyurethane (PU) foam for domestic within the main application H andrefrigerators for(the home marketd subapplication S. Demand is measured in

refrigerators for the home market; ozone-depleting potential (ODP) weightede PU foam for domestic refrigerators for tons.

export; Projections are then made on the basis of

* other Pu insulation foam; and factors which are expected to influence thedemand for products that contain or are made

* extruded polystyrene packaging foam. with ODS, for instance

In the Tunisia Country Report no * population growth rate;

subapplications were defined for the majoruser sectors: flexible foam, aerosols, solvents, * income growth rate;

and fire extinguishants (H 3 -H 6 ). For * sector plans (near-term expectations of

Czechoslovakia, the solvents sector was industry leaders);

divided in four subapplications: electronics,

dry cleaning, metal degreasing, and other * present state of market saturation for

applications. consumer goods, for instance, refrigerators,

The reason for distinguishing between aerosol products, and mattresses; and

domestic refrigeration for the home market * the quality of consumer products (service

and for export is that the 10-year delay needs and standards).

provision in the Protocol only applies to the


A view is also made to the past The control options are denoted by Cj in thedevelopment in ODS consumption. Table 5-2 cFccost model. For example, control optionsbelow summarizes the basic assumptions in the commercial refrigeration sector may bemade for Tunisia concerning the growth rateof individual ODS applications during the H,S4C1: better maintenance and servicingperiod 1991-2010. The future demand (in year procedures (a process substitute);t) is determined on the basis of DO(H,S) and H, 4C,: HFC-134a (a chemical substitute);the expected growth rate V(H,S): and

H1S4C3: HCFC-22 (a transitional chemicalD(H,S,t) = DO(H,S) * ( +V(H,S))', t = I,..., t; substitute).

1 = 1992; T =2010.

The model assumes that the control optionsThe model allows the user to specify are independent of each other so that

varying growth rates over the period of reductions attainable from each control can beanalysis; for example, ODS demand in many added. For each control option, the followingdeveloping countries may be expected to grow input is specified:at a faster rate in the beginning of the period * assumed earliest start date (denoted bythan towards the end of the period. BEGIN[H,S,C]), which reflects the year in

which a control is expected to beStep 3: Technical assessment of the most commercially available in Tunisia orappropriate substitution techniques for each Czechoslovakia;ODS application

* market penetration time (denoted byFor each specific ODS application (HoSn), PEN[H,S,C]), which is an estimate of how

feasible alternative technologies (or control long it will take for a control to fullyoptions) are specified. Only the most plausible penetrate the market to reach the maximumalternatives to each subapplication are reduction potential;included. Examples of alternativetechnologies are: reduction potential (denoted by

PRO[H,S,C]), that is, the maximum* Product substitutes, that is, replacement of percentage reduction in ODS use within a

an ODS-consuming product by a non- (or subapplication that can be expected whenless) CFC-producing product such as roll-on the control has fully penetrated the market;deodorants, paper, and cardboard andpackaging;

* unit incremental substitution costs (see* Process substitutes, for instance, recovery Step 4 below).

and recycling and mechanical pumps; andThere is an interrelationship between unit

* Chlemnical substitutes, that is, use of a non- substitution costs on the one hand and earliestor less ozone-depleting chemical than the start date and market penetration time on thecontrolled substance or change in the other. It may be feasible to implement aproportions by which CFC and other control earlier and/or with shorter penetrationchemicals enter into a traditional mixture, time than stated in the model, but that wouldfor example, hydrofluorocarbons (HFCs), generally imply higher substitution costs.hydrochlorofluorocarbons (HCFCs), liquidpetroleum gas (LPG), or 50 percent CFC

reduced foam (reduced with water).


Step 4: Estimation of the incremental user include the potential economic savings whichcosts of each replacemlent techinique could be gained by closing down small-scale,

Icmtlosadffet inefficient production in connection with theIncremental transition to ODS substitution technologies.

The incremental user costs include:ODS producer (Czechoslovakia); * capital cost of new production facilities or

* ODS user industries; conversion of existing facilities and of new

* at the level of final consumers (see Step 5); equipment, such as recovery and recyclingand machines;

* in the government sector. * cost of patents and designs and incrementalcost of royalties;

The general principle is to identify costswhere they first arise, with one exception: we * operational costs, including raw materials,assume that the costs arising in the producer labor, energy, and waste disposal;sector (incremental costs of building or * costs of retraining personnel, includingmodifying a CFC plant to make a chemical research and development to adaptsubstitute) are passed on to the users in the technology to local circumstances; andform of higher prices on ODS substitutes. Thisexception from the general principle is made stpoftein order to avoid double counting, and in order support.to include incremental costs of raw materials The unit incremental substitution costs,in the cost calculations for countries that rely measured in US$ per kilogram of ODS useon ODS imports. avoided, are calculated as follows:

The incremental costs to the country are COST CC * a + NOCcalculated as economic costs rather than QODSfinancial costs; that is, transfer payments where:between the public and the private sector are COST : unit substitution costnot included. Therefore, unless special local CC : incremental capital and other one-conditions call for adjustments in the cost time costs such as training,levels (for instance, in the case of labor costs), technical assistance, and so onwe have used world market prices on inputs a : annualization factor, defined as(for example, on materials and machinery). a = rFreight and other related costs are only 1 - [1/(l+r)']assessed in cases where ODS imports arereplaced by locally-produced substitutes (as in r : real discount ratethe case of LPG for aerosols in Tunisia, where NOCS ntincremental er coststhe incremental cost is estimated as the QODS : quantity of ODS saved per year ifdifference between the local ODS the control option is implemented.price-excluding taxes and the local price of The government itself will incur economicLPG-implicitly taking account of the freight costs in complying with the Montrealcosts). Otherwise, we assume that freight costs Protocol, for instance, for implementing thewili be the same in the with and without necessary laws and regulations, monitoring thescenarios. use of ODS, and disseminating information.

Also, it should be noted that all incremental All of these government costs are regarded ascosts are calculated on the basis of existing incremental, but they are not included in theeconomic and industrial policies in the model calculations because they do not affectcountry (Tunisia). Thus, the model does not the choice of phaseout strategy.


Step 5: Estimation of the incremental necessary to calculate the incremental costsconsumer costs ofpremature obsolescence and optimum phaseout schedules for eachof refrigeration equipment group separately, In other words, one would

The model calculates the costs incurred need four models in this case. However, thisdirectly by the final consumers resulting from distinction has not been made in the Tunisiaforced early replacement of refrigerators Country Report because halons, MCF, and CTC

forcd eonly represent 2 percent of the total ODPwhich cannot be recharged because of weighted ODS consumption. In other words, itnonavailability of ODS. (It is assumed that is assumed that all ODSs are regulated by theCFC- 12 is available for servicing up to the year same phaseout requirements as CFCs.2010 but that none will be available thereafter. According to the London Amendments, theseThe option of recycling is discussed and phaseout requirements include:analyzed as part of the sensitivity analyses.)

It is assumed that the new refrigerant will * a freeze by 1999 compared to base yearbe as energy efficient as the CFC-based ones. consumption;

Possible forced early replacement costs of * a 50 percent reduction by 2005;other commercial and industrial refrigerationand air conditioning equipment and fire * an 85 percentreductionby 2007; andfighting installations are not included in the * a 100 percent reduction by 2010.model, partly because it is assumed that most According to the Protocol, the base againstof these can be retrofitted, and partly because which ODS reduction requirements arenot enough data is available. measured is defined as the average

The costs to consumers resulting from consumption in 1995, 1996, and 1997 or 0.3higher purchasing prices of non-ODS-based kg per capita if this were lower. In Tunisia, theproducts are not included as consumer costs, projected unconstrained demand for ODS inas this would involve double counting. In the years 1995-1997 is well below the 0.3 kgorder to calculate the incremental consumer limit, so the first mentioned base definitioncosts, the following data are needed: applies. In the model, the constraints imposed* the present stock and age profile of by the Montreal Protocol are denoted by

refrigerators; MON(t) and the base year ODS consumption

- the annual production of refrigerators; by BASIC.The technical and institutional constraints

* the rate of servicing (recharging) of are given by the start date, market penetrationrefrigerators; and times, and reduction potentials of each control

* the present price, excluding taxes, of a new option.refrigerator (average).

Step 7: Specification of alternative ODSStep 6: Specification of the constraints puhaseout strategiesimposed by the Montreal Protocol as well aspotential technical and institutional Three alternative phaseout scenarios wereconstraints analyzed for Tunisia:

The Montreal Protocol imposes different * Allowable phaseout-a scenario thatphaseout requirements for different groups of postpones ODS phaseout until the latestODSs (CFCs, halons, methyl chloroform IMCF], possible date while keeping the countrycarbon tetrachloride [CTC]), and substitution within the limits of the Montreal Protocol.between groups of ODSs is not credited for * Phiaseout at minimum incremental costscompliance purposes. Thus, in principle, it is (least-cost scenario).


* Accelerated phaseout-phaseout at a pace choice of available substitutionset by the availability of technology, technologies;(option of lowest ODs emissions). * the cost-effective total substitution costs forThe first and second scenarios are the individual application areas (user

characterized by full utilization of the 10-year costs);delay provision in the Protocol (except for the costs to the consumers of prematuredomestic refrigerators in the second scenario). retiring of refrigerators.Under the last scenario, the transition to ODSsubstitutes is accomplished over the period Step 9: Cashflow analysis1992-1999, except for servicing ofrefrigeration equipment, which is allowed up Following the estimation of the most cost-to year 2010. The reduction schedule for the effective choice of control options and timingthird scenario is based on a model calculation of phaseout in the individual user sectors, aof the maximum feasible reduction in percent cash flow is calculated for each scenario.of initial ODS demand. For Tunisia, the Costs are again divided into the followingscenario runs thus: categories for each subapplication:Base year 1991 * incremental capital and other one-time

a Freeze by 1993; costs;

= 50 percent by 1995; * net incremental operating costs; and

= 85 percent by 1997; * incremental consumer costs of premature

* 90 percent by 1999; retiring of refrigerators.- 95 percent by 2005; and The timing of investments is determined

* 95 percent by2005;andaccording to the optimum phaseout schedule* 100 percent by 2010. resulting from Step 8.

Step 8: Calculation (minimization) of the Step 10: Sensitivity analysesincremental costs of each phaseout scenariorelative to the without protocol situation The calculation of incremental costs of

alternative phaseout scenarios is supplementedThe total incremental costs are calculated by analyses of sensitivity with respect to:

(minimized) subject to the constraints of each . .phaseout scenario. The model operations aredescribed in the section Description of Model * assumed servicing rates for domesticfor Calculating Incremental Costs of ODS refrigerators; andPhaseout and result in the calculation of the * assumed growth rates in unconstrainedfollowing output: demand.

* the present value of total incremental cost It should be noted that analysis of theof ODS substitution during the period sensitivity of the results to changed ODS pfices1991-2010; the cost is expressed in 1991 is the same as analysis of changed substitutionU.S. dollars; costs: higher ODS prices mean lower

incremental unit substitution costs. Thei the total accumulated ODS use (1991-2010) sensitivity analyses are presented in a sectionin the with and without situation; below.

* the optimum choice of control options, thatis, the most cost-effective timing and


Technical and Economic Assumptions applications and many small users. This is

The technical and economic assumptions particularly the case in the solventsThe ~ ~ ~ ~ ~ ~ sctrue ofhnca Mnd andoml CTCumtin

made in the Country Program for Tunisia are sector-uses of MCF and CTC wn

summarized in Tables 5-2 and 5-3. A real Czechoslovakia, for example, are not welldiscount rate of 10 percent is used in all model established-but also in part of the

refrigeration sector, such as in the servicing ofcalculations toae capital cost overe domestic appliances, as well as in commercial

cash flows to iem ent a i nue and transport refrigeration (for instance,refrigerated fishing vessels and other vessels

Special Conceptual Features and with refrigeration equipment on board).

Problems Step 2: The problem of defining the without

For each step, 1 to 7 above, this section scenario

discusses any special conceptual features, The projection of unconstrained demanddifficulties in obtaining relevant data, or

assumes that no countries had imposed theproblems of interpretation in the country. Protocol, including its restrictions on trade.

This implies that losses connected to trade

Sonsumption data with developed countries that have ratified theProtocol are regarded as incremental and

Difficulties were experienced both in therefore could be compensated by theobtaining data for the total consumption Multilateral Fund or the Global Environment(= production + import - export) of individual Facility (GEF), although trade losses aresubstances (CFCs, halons, crc, MCF and HCFCs) normally not compensated.and information about the distribution on Furthermore, any future increases in thespecific applications. real prices of ODS are regarded as incremental

In Tunisia, import statistics are not easily costs, since we assume continued availabilityaccessible or catalogued with monitoring of at current prices. Future increases in the realODS use in mind; also, the statistics might be prices of ODS could result from factors otherincomplete, and they do not identify ODS by than the Protocol; the current ODS price couldtype. Interviews with the key chemical already be influenced by the Protocol as inimporters and distributors only provide part of Europe now because of an oversupply of CFC.the picture because several of the larger users Therefore, in order to avoid overcom-have their own direct import of ODS. Data pensation, it is important that the incrementalcollection, therefore, focused on interviews costs of alternative substances like HFCs andwith the user industries. the transitional HCFCs are calculated on the

In Czechoslovakia, user surveys covering basis of equilibrium prices, that is, prices that90 percent of consumption have been carried are expected to prevail once the ODSout since 1989. There is one state-owned replacement has been made.producer and until recently only one importer, Unconstrained demand is projected up toso the total consumption is fairly easy to the year 2010 both for Tunisia andobtain. However, as new private companies Czechoslovakia. The calculation of

emerge and customs control is liberalized, the incremental user costs thus assumes that costsmonitoring system will have to be extended. are compensable over this time horizon both

The assessment of the distribution of ODS from the Multilateral Fund and from the GEF.

use on specific applications is a difficult or at Another problem arises already whenleast time-consuming task in all countries. In determining the base year ODS consumption.

some user sectors, there are numerous minor First of all, it must be adjusted for already

5. Evaluation of Incremental Cost Methodologies for oDS Phaseout Case Studies 91

substituted ODS as a result of the Montreal situation makes it difficult to choose the mostProtocol, that is, after 1987. Secondly, it plausible combination of control options.should be adjusted for temporary plant close- Uncertainty also exists as to the earliest startdowns, as has been observed in date, market penetration times, and reductionCzechoslovakia and other Central and Eastern potentials, which may vary from one countryEuropean countries since the breakdown of to another depending on institutional capacity,the former centrally-planned economies. This economic policy, and the general economicis necessary in order to make sure that structure. Therefore, a lot of best guesses, orcountries that have already taken steps to qualified guesses, have to be made based oncomply with the Protocol are not available data and experience from otherdisadvantaged vis-a-vis countries where no countries.actions have been taken yet.

It is assumed that higher- or lower-priced Step 4: problems of estimating incrementalalternatives to ODS do not affect the demand substitution costsfor ODS-based products. This is justified ifincremental costs are compensated and not large t ec tong the os of

pased n t th cosumrs,butif naretsare large but unknown effect on the costs ofpassed on to the consumers, but if markets are supplying these techniques as well as on thenot competitive, this will seldom be the case iuced tst tthe csumers If or

' . . ~~~~~induced costs to the consumers. If, forin practice. On the other hand, too little is instance, success is made in finding anknown about the elasticities of demand for acceptable drop-in substitutelfor CFC-12 inODS-based products to take account of this domestic refrigerators, consumers will notinformation In the demand projections.

Becauseaverynlittlehinformationpisjgenerally suffer any forced early replacement costs, andavcailabe oncernlinthle ifaorsathat influenery the cost-effective phaseout strategy may beavalable conceem'ng the factors that Influence

aval I completely altered. Another example is thethe demiand for ODS-based products, it is not uncertainty related to the assessment of thepossible to make scientific projections of substitution costs in the commercialuinconstrained demand-certainly not up to refi-igeration sector, where each case has to bethe year 2010. Therefore, the projections made evaluated individually. It may be possible tofor both Tunisia and Czechoslovakia should retrofit one installation but not anotherbe interpreted as scenalrios rather than as

actual forecasts. ~~~without significant extra energy costs oractual forecasts...modification costs.

Stel) 3: uncertainty concerning tcclitiictThe social costs of potential changes in the

alterLativ es quality of the product as a result ofsubstitution are not included in the

The search for suitable alternatives to ODS incremental cost calculations.has been and still is extensive; a wide range of The principle of identifying costs wherefeasible alternatives exists for each ODS they first arise (except in the ODS producerapplication. These alternatives may or, as in sector) is problematic if the country does notmost cases, may not have exactly the same manufacture its own ODS-based equipmentproperties as the well-known and tested ODS. (refrigerators, for example) but relies onOne alternative may be the most likely today import, often of second hand, equipment. Forbut become unlikely tomorrow when the instance, in the case of Ghana, we includedresults of new tests become available. the extra costs of the ODS-free refrigerators inDevelopment has been very fast in this field. the country's incremental consumer costs.Furthermore, one firm or country may choose This problem also arises in connection withone alternative, whereas this may not be the assessment of import of intermediary oracceptable to another firm or country. This final produLcts containing ODS, for instance,


halon import in portable fire extinguishers heavy up-front investment costs which cannotfrom the former German Democratic Republic be met because of capital constraints.to the Czech and Slovak Federal Republic One of the largest uncertainties in the(CSFR) or Bulgaria's premixed foam import calculation of unit substitution costs arisesfrom Italy. The ODS content should be when trying to estimate the amount of ODSregistered as consumption in Germany and that can be saved from investing in theItaly, but both Czechoslovakia and Bulgaria alternative technology. The capacityhave the problem of retrofitting, retiring, or utilization of the equipment often determinesconverting production facilities to alternative whether the technique is cost-effective or not.technologies. Therefore, incremental costs For example, a recycling machine used in awould be underestimated if these uses were large repair shop with centralized repairs maynot included. be compared to the same machine used by a

By using world market prices on major small service shop for on-site repairs. Thus,inputs such as raw materials, energy, and depending on assumptions about capacitymachinery, we implicitly assume that the utilization, unit costs may vary considerablyeconomy is open. This is a reasonable long- from country to country, even though capitalterm assumption. Although market and operating costs are almost the same.imperfections certainly exist, especially in theshort term in the former Council for Mutual Step 5: problems of estimating costs ofEconomic Assistance (CMEA) countries, it forced early replacementwould not be justifiable to spend resourcestrying to adjust the costs accordingly. Those The es tes ed earlytreplacemencosts are extremely sensitive to thecosts will certainly change over time, and assumptions made about the future availabilityincremental user costs are already surrounded of drop-in substitutes as well as available ODSby enough uncertainty. om rop-in and rell as important,

The incremental cost estimates obtained from recycling and reclaiming. It is important,from the industries should be critically therefore, that these key assumptions bereviewed, as their prices tend to be on the high included in the sensitivity analyses.side. Perhaps they are relevant as short-term Only forced early replacement costs ofcosts, but most often these are too high domestic refrigerators are included. If other

compared to long-tl. ODS-based equipment that requires ODS forcompared to long-term equilibrium prices. se.cn weeicue,dtapolm.olFurthermore, care should be taken to adjust servcing were Included, data problems wouldfinancial costs where these differ from be insurmountable. This is true not only

because the stock of other ODS-basedeconomic costs. In terms of compensation equipment is not known beforehand, but

from the Multilateral Fund and the GEF, this because the price and size of this equipmentadjustment raises a problem, because in has extreme variations. There is a dearth ofpractice, compensation is likely to be for knowledge about how many of thesefinancial costs, since it is to these costs that in ons about how many

the~ ~ usridsre. n cnuesrsod installations can be retrofitted, and how manythe user industries and consumers respond... The Fund, in particular, focuses on grant require significant modifications or completefinancing of specific phaseout projects at the replacement.industry level. To maximize the benefits from The calculation of forced early replacementthe Fund, there is a need for a more flexible costs of domestic refrigerators is morefinancial mechanism (loan and revolving fund complicated in countries where no productiontype). The aerosol sector is an example: net of refrigerators takes place and a substantialincremental costs are negative but there are share of the refrigerators are secondhand. The

used refrigerators are much cheaper than thenew ones but consume much more energy. It

5. Evaluation of Incremental Cost Metjiodologies for ODS Phaseout Case Studies 93

is not known whether the secondhand Program preparation (analyses of alternativerefrigerators will be replaced by new or used phaseout scenarios) may be questioned in theODS-free refrigerators when CFC is no longer case of these countries. This point is discussedavailable. These considerations have to be in further detail in the section Drawingincorporated into the calculation of consumer Lessons in Methodology.incremental costs in countries that rely heavilyon secondhand imports. There are also Model for Calculating Incremental Costssignificant distributional problems in of ODS Phaseoutcountries where the price of electricity is The CFC CoSt model identifies the cost-

subsidized. minimizing cornbination and timing of

Step 6: different phaseout requirements for phaseout actions subject to the requirementsdifferent 'O'S'of the Montreal Protocol and the technical and

diff$erent ODSsinstitutional constraints. It is a cost-

This problem was discussed earlier. In effectiveness model rather than a cost-benefitcases where ODSs other than CFCs constitute a model: it identifies the most cost-effectivesignificant share of total ODS consumption, way of reaching a given environmental targetone model or incremental cost calculation (the Montreal Protocol or alternative reductionshould be made for each significant group of requirements), but is not able to determinesubstances. whether the reduction requirements

themselves are optimal from a cost-benefitStep 7: alternative ODS phaseout strategies point of view. The benefits are calculated in

For CEE countries and those cis countries terms of ODS use avoided by implementing thethat will not obtain the status of developing cost-effective phaseout strategy, but the

countr, te abenefits are not valuated in monetary terms.countries, the actual difference in alternative Temdli optrbsdlnaphaseout scenarios has all but completely The model s a computer-based lGneardisappeared due to the November 1992 prograimming model using the General

tightning f th Monteal rotocl in Algebraic Modelling System (GAMS). It can betighengng ofthe Montreal Prtoo in executed on a PC with at least a 386 processor

Copenhagen.~~ ~ WihJnay1,19,a' h and a minimum of two megabytes of RAM.required phaseout date for CFCs, MCF, and CTC T bic optimizatio ob isand January 1, 1994, for halons, there are few Te as ollws(crrson to thoptions but to shift to non-ODS technologies as leat-o scao):soon as technically and institutionally feasible.Therefore, the general framework for Country

94 OZONE LAYER PROTECTION: COUNTRY INCRENENTAL COSTS

minZ = * X(H,S,C,t) * D(H,S,t) * COST(H,S,C)/1000 + S OLD(v,2010) * SCRAPCOST(v)x +r) 0c1 +rI

s.t. (1) X(H,S,C,t) s REDUC(H,S,C,t) H,S,C,t

(2) X(H,S,C,t) = REDUC(H,S,C,t) if COST(H,S,C) s 0 H,S,C,t

(3) F X(H,S,C,t) s I H,S,tC

(4) _ DO(H,S) * (I -MON(t)) ( - E SHARE(H,S) * X(H,S,C,t)) * ZD(H,S,t) t1i.5 1H S,C H,S

(5) NEW(l,t) + OLD(1,t) s REF * D(refrigeration,domestic,t) t

(6) NEW(v + I,t + I) = NEW(v,t) v,tOLD(v + I,t + 1) = OLD(v,t) v,t

(7) NEW( I ,t) = REF * X(refrigeration,domestic,t) * D(refrigeration,domestic,t) t

(8) SERVICE(t) = F OLD(v,t) * SERV(v) * CFCSERV t

(9) Y X(refrigeration,servicing,C,t) * D(refrigeration,servicing,t) s SERVICE(t)

where

* X is the decision variable and indicates the reduction achieved for each control option-in percent of theprojected ODS demand for the corresponding subapplication;

- r is the discount rate used to discount the costs to the base year; and

* SCRAPCOST (v) is defined as

21-Y NEWPRICE * SCRAPSHARE(i) NEWvPRICE

Li=, (I~~+r)'~ J (r)2

where

v refrigerator vintage, v = ],...,20

NEWPRICE price of a new CFC-free refrigeratorSCRAPSHARE share of refrigerators that have to be retired in each vintage because no CFC is

available for servicing


The parameter REDUC indicates for every X*D*COST*1000 indicates for every sub-control option the technically feasible application the eliminated CFC use inreduction potential in percent of the kilograms (D is expressed in ODP tons)projected ODS demand for the multiplied by the unit substitution cost. Bycorresponding subapplication. When the totaling all subapplications, the total usercontrol has fully penetrated the market, substitution cost is calculated. To this must beREDUC is equal to PRO(H,S,c). The added the costs to the consumers of prematurereduction potential for the preceding years retiring of refrigerators, OLD*SCRAPCOST foris approximated by using linear every refrigerator vintage from 1 to 20 (theinterpolation. assumed lifetime of a refrigerator).

* SHARE(H,S) indicates for every The constraints (1), (2), and (3) define thesubapplication the share of total ODS technically feasible solutions for every controldemand. If the growth rates, used to project option. From constraint (1) it is evident thatthe future unregulated demand, differ the control options are assumed to beacross applications, the SHARE parameter independent (the reduction potential for onewill not remain constant. In that case, control does not depend upon the level ofSHARE must be a function of t as well as of other control options chosen). Constraint (2)H and S. ensures that ODS reduction is maximized once

costs are minimized., that is, that control* MON(t) indicates for every year the options with zero substitution costs are

reduction requirements imposed. MON is implemented as soon as technically feasible.expressed in percent of the total ODS Constraint (3) ensures that the reductiondemand in the base year. obtained for each subapplication does not

* NEW(v, t) and OLD(V, t) are the number of exceed 100 percent.HFC-134a and cFc-12 based refrigerators, Constraint (4) reflects the reductionrespectively, for each refrigerator vintage requirements. The left-hand side gives theand every year. number of ODP tons of ODS use remaining

after the cost-minimizing substitution has• REF is the number of refrigerators which taken place, and the right-hand side indicates

can be produced from I ton of CFC- 12; how many ODP tons of ODS use are allowed to

* SERVICE(t) is the demand for CFC for remain when the reduction requirements mustservicing of domestic refrigerators in each be fulfilled.year. Constraints (5) to (9) are all related to the

* SERV(v) is the recharge requirement for calculation of the forced early replacementeach refrigerator vintage (measured in costs of refrigerators.percent of a vintage). Analysis of the allowable phaseout scenario

is based on the same model input as the least-uFsedV isc thmeoneavefrageramount of seicFe cost scenario; account is not taken of theused each time one refrigeratoris serviced forced early replacement costs when(for purging, leak testing, recharging, and determining the optimum choice and timing ofso on). phaseout actions. In other words, the

The problem is to choose the control consumer costs are not included in the costoptions in such a way that the total discounted minimization problem but are added to totalincremental economic cost Z is minimized, costs afterwards.while at the same time fulfilling the reduction Finally, the accelerated phaseout is basedrequirements. Z is expressed in millions of on the same optimization routine as the least-U.S. dollars. The objective function cost (that is, forced early replacement costs are


included), but the input phaseout requirements Net present value capital, operating, andMON(t) are changed to reflect the maximum forced early replacement costs for eachfeasible reduction given by the technical and scenario are shown in Table 5-5. Theinstitutional constraints. incremental annual cash flows in fixed prices

from 1991 to 2015 are illustrated in FiguresSummary of Country Incremental Costs 5-2 and 5-3 for the allowable and accelerated

The model calculations result in optimum phaseout scenarios.The capital costs of ODS phaseout only

1991consumtion phofis over thre phseriod constitute a small share of total phaseout199 1-2010 in each of the three phaseout css nrmna prtn oti hstrateges show in Figre 5-1.costs. Incremental operating cost is thestrategies shown in Figure 5-1. dominant cost element in Scenario 2 and 3 in

In the first strategy, the production of -1 134a Tal5-dutohexecdhieriesf' ~~~~~~~Table 5-5 due to the expected higher prices ofrefrigerators for the local market is postponed intermediate components and HFC and

until 2009, just in time for a complete transitional HCFC raw materials. The costs tophaseout in 2010. In scenario two, production the consumers of forced early replacement ofof cFc-free refrigerators is initiated as early as refrigerators is the dominant cost element inpossible, that is, by 1996, in order to minimize the allowable haseout scenario.total incremental costs to the Tunisianeconomy. Sensitivity of the Cost Estimates

Although the consumption profiles in thefirst two scenarios are very similar, the Based on analyses of the sensitivity of thedifference in total incremental costs is model results to changes in the criticalsignificant due to the fact that the total costs in assumptions (ODS prices, ODS availability,the allowable strategy are dominated by the servicing rates for refrigerators andcosts associated with forced early replacement unconstrained ODS demand growth rates), theof domestic refrigerators, as shown in Table following can be concluded:5-4. Minimum net incremental phaseout costs * A doubling of the present ODS prices willare estimated at about $25 million discounted make most of the technical substitutionto 1991 values compared to approximately options economically attractive, whereby$100 million in the allowable phaseout the accelerated phaseout scenario will bescenario. identical to the least-cost scenario;

It is assumed that economically viablephaseout actions (net incremental costs are * If CFCs are no longer available for Tunisiazero or negative) will be implemented to import after the year 2000 and no drop-inregardless of Protocol restrictions. This is the substitutes become available, totalreason the consumption profiles do not follow incremental costs will increase significantlythe Protocol limits in the beginning. from about $30 million to $100 million,

The extra costs to Tunisia of implementing even if the accelerated phaseout schedule isthe accelerated phaseout (strategy 3) compared followed and the production of cFc-basedto the minimum cost strategy are estimated at refrigerators is stopped by 1996. In thisless than $5 million. Conversely, the extra case, a comprehensive recovery, recyclingenvironmental benefits in terms of ODS and reclaiming scheme for Tunisia isconsumption avoided are considerable: 12,000 economically feasible. Recycled CFCs

ODP tons. Thus, the marginal reduction cost of would serve as a supply source reducing theimplementing the accelerated phaseout instead need to retire refrigerators before the end ofof the minimum cost phaseout is only $0.3/kg their lifetime.of avoided ODS emissions. * In Czechoslovakia, where new CFC will be

banned from January 1, 1996, forced early


replacement costs have been estimated at industries and consumers of the Montrealmore than $150 million in present value Protocol. Furthermore, an acceleratedterms. These costs can only be avoided if CFC phaseout is in accordance with the projectsis hoarded in the coming years, CFC can be proposed for financing from the Multilateralsupplied from recycling, or a drop-in Fund and in agreement with the aim of thesubstitute becomes available. Fund to provide incentives for early action.

• The results are not sensitive to changed The reasons this phaseout strategy isassumptions about the rate of servicing of recommended as opposed to the least-costdomestic refrigerators. Even if failure rates phaseout schedule are the following:are much lower than experienced in Tunisia * First, although the phaseout schedule is nottoday, the conclusion that an accelerated the cheapest, it is only slightly morephaseout of CFCs in production of new expensive than the least-cost scenario,refrigerators is the most cost-effective does whereas the additional environmentalnot change. benefits are significant.

* A 50 percent increase (decrease) in * Second, if CFC prices rise, as can beassumed growth rates of unconstrained expected from 1995 or earlier when ODS

demand increases (decreases) total production will be banned in the Europeanincremental phaseout costs by 25-30 Community, substitution costs will in manypercent. cases become negative, implying that an

The sensitivity of the discount rate was not early phaseout is economicallyanalyzed in the Tunisia and Czechoslovakia advantageous.Country Program Reports. However, earlier Furthermore, some of the major users inanalyses of ODS phaseout strategies in the Tunisia have already taken actions toPhilippines indicate that the discount rate has eliminate their use of ODS, or have formulateda significant effect on the absolute net present plans to do so despite incurring extra costs.value of incremental costs. It does not, This is either because expatriate businesshowever, affect the optimum choice of owners or partners are under pressure fromphaseout strategy. consumer groups abroad, or because ODS-

based export products must be replaced by

Recommended Policies 1996 (the phaseout date in developedcountries). As a result, it may be cost-effective

Overall ODS Phaseout Strategy and to switch the whole production in one step,Policy Instruments even though only part of the production is

We recommend that the Tunisian exported.government adopt an ODS phaseout schedule The total incremental costs of thethat corresponds to the accelerated phaseout recommended phaseout schedule is estimatedscenario, This phaseout schedule proposes a at $29 million (net present value). The benefitfereeze This phaseoution scoedule proposes1 is a reduction in total unconstrained ODSfreeze in consumption compared to 1991 consumption of almost 85 percent over thelevels by 1993, a 50 percent reduction by cosmtnofaot85pretvrth1995, 85 percent reduction by 1997, 90 period 1991-2010. The planned consumptionpercent reduction by 1999, and 100 percent of ODSs until complete phaseout is shown inreduction by 2010. This schedule is well Figure 5-4 by end user sector.ahead of Protocol limits and reflects the In addition to the overall phaseout strategy,government's objectives to minimize an Action Plan covering an initial period of

Tunisia's consumption of ODS while at the three years (1993-1995) to implement thesame time minimizing the cost to Tunisian strategy is recommended. The plan comprises

98 OZONE LAYER PROTECTION: COUNTRY INCRENENTAL COSTS

the introduction of government initiatives to regulations, new developments on theeffectively regulate and monitor the use of markets for ODS substitutes, and the resultsODS, and a series of proposed projects in the of the proposed demonstration projects.ODS end user sectors, all of which are seen as From an economic efficiency point of view,prerequisites for implementation of the it is generally recommended to control ODS

phaseout schedule. supplies rather than ODS use, and leave it toA summary of recommended government the market to allocate scarce ODS supplies

actions or policy instruments is shown in among the users. However, under such aTable 5-6. These comprise: system there will also be a tendency to utilize

* Creation of an Ozone Committee for the maximum allowable ODS consumption;implementation of the Action Plan. The where market distortions exist or fulloverall objectives of the committee should information is not available, this system willb e to initiate and facilitate the disfavor the small user industries. It is,implementation of the Action Plan. It is therefore, strongly recommended toenvisaged that the committee will be supplement supply control by user specificsupported by an ODS working group with bans and control. Doing so would also preventrepresentatives from each of the major users from replacing ODS by alternativeindustry sectors using ODS. A full time technologies that may have adverse humancoordinator of the committee is proposed. health effects.

* Setting up a monitoring system for ODS use The applicability of the above regulatoryby substance and end user sector. An instruments in Central and Eastern Europe hasimport license will enable the government been studied in a recent report on regionalto effectively control the supply of ODS. considerations for the phasing out of ODS inFurthermore, ODS users should be obliged CEE countries (Commission of the Europeanto report annually their purchase and Communities and the World Bank 1993).consumption of ODS so that considerate zere,thepolicyinstrumentsrecommendedtforpolicy actions and adjustments of the Czechoslovakia and other CEE countriesAction Plan can be made if necessary. include the following:

* Regulatory measures which include sector- * A tax on the production and import of ODSspecific bans on the use of ODS; a special is recommended. First, an ODS tax mayODS tax on imports of all controlled provide an important incentive to recyclesubstances to encourage initial substitution ODS in order to avoid high costs of forcedof ODS in easily substitutable applications early retirement of ODS-based equipment.such as aerosols and foams; tax exemptions Furthermore, the fiscal effect of the tax isfor imports of ODS-conserving technology; particularly important in transitionaland standards for recycled CFC and flexible economies where other revenue raisingfoam densities. instruments, such as corporate and income

taxes, are not yet well developed and where* Introduction of an accreditation system for public companies still play a dominant role.

refrigeration technicians with the purpose It is emphasized, however, that the generalof supporting and enforcing ODS incentive effect of a tax is probably limitedconservation methods during service of due to inflation, soft budget constraints,refrigeration equipment. and uncertainty. A tax or other economic

* Information dissemination with the purpose instrument should not stand alone butof increasing consumer awareness of the should be used in combination withozone problem and informing the ODS user administrative regulation, such as importindustry about the most recent Protocol licenses and sector specific bans. This is


particularly important in transitional development of monitoring andeconomies due to the predominant market enforcement measures to supportimperfections. The distribution of govern- regulatory measures;ment action costs on individual items is * modification of certain productarbitrary, as the Ozone Committee secretary standards and norms in support of thewill be involved in all activities. introduction of non-ODS alternatives in

* Tradable permits are not recommended as some applications; andpart of the ODS regulation in CEE countries, * extensive efforts in awareness-buildingmainly due to the difficulties in in the form of information disseminationdeterrmining the initial allocation of permits and product labeling.in a situation where old industrial and tradestructures are changing rapidly and due to The type of ODS phaseout projects proposedthe risk of creating large quasi-rents in the for Tunisia for funding from the Multilateralcase of auctioning. Fund are shown in the project summary figure

in Table 5-6. The type of projects recom-* Economic support (for instance, for mended for each user sector are almostresearch and development projects or generic, whereas the size of the projects (totalinvestment in ODS free technology) is costs and benefits), organization, and detailedrecommended in order to assist ODS design are country specific.industries in CEE countries in complyingwith the tight phaseout schedule of theCopenhagen Amendments. Furthermore, there is a need for establishing a financial Pros and Cons of the CFC Costmechanism for ODS reduction. Optimization Model and the Need for

* Both sector-specific bans and production Model and Scenario Modificationsand import licenses are recom- The main strength of the CFC costmended-sector specific bans in order to optimization model is that it automaticallyallow the most essential uses (refrigeration selects the scenario with the lowestand some solvents and rigid foam incremental costs subject to the Montrealapplications) to be maintained until the Protocol requirements and the specifiedlatest possible date, and production and technical and institutional constraints. Trial

imiport licenses in order to facilitate and error methods of finding the least-costmiionitoring and control. scenario are not necessary. The results of the

* In addition to an appropriate mix of model have pointed out that the acceleratedadministrative measures and economic phaseout schedule is not the least-costinstruments, a comprehensive ODS phaseout schedule, except in the case of domesticstrategy requires the following supportive refrigeration and possibly some other sectorsmcasures: that may endure premature retiring costs. Only

introduction of emission control if we assume that ODS prices rise during themieasures to conserve the large stock of phaseout period will the accelerated pliaseoutODS in refrigeration and cooling schedule be identical to the least-cost. Apartequipment which can be reused to from the environmental argument, rising ODSextend the economic life of these prices are the main reason for recommendingequipment the accelerated phaseout.

Presently, the price issue is analyzed as partof the sensitivity analyses; it would increasethe validity of the model if price changes over


the period of analysis were incorporated countries because of the present capitaldirectly. This would be fairly easy to do from constraint (financial market imperfections),a modeling viewpoint. The problem lies on the and should therefore play a key role whendata side. choosing between the few available phaseout

Another element which needs to be options and deciding on project priorities.included more directly in the model analyses Of course, analyses of alternative phaseoutis the supply side, which is extremely relevant scenarios still make sense in the case ofin connection with the calculation of forced developing countries because of the 10-yearearly replacement costs and subsequent delay provision. However, strategies, policyrecommendations made for phaseout projects. instruments, and the type of projects beingAlternative strategies or options exist with proposed tend to become generic, as describedregard to the future supply of ODS: recycling below.and reclaiming, stockpiling ODS for futureservicing needs, continued ODS SUpply to General Policy Research and Datadeveloping countries from chemical producers The costs of ODS substitute technologies

up to the year 2010, and discovery of drop-in tend to become generic if world market pricessubstitutes. These options need to be analyzed are used, as can be argued for openin further detail, not only at a country level, economies. Therefore, a worldwide applicablebut also on an overall strategy level. but asoanor r. cost catalogue should be developed which can

The CFC cost model was originally be modified according to special localdeveloped to analyze alternative phaseout conditions (where these modifications can bestrategies at a time when the Montreal .f.Protocol only demanded a 50 percent justified).

At present, incremental cost estimates madereduction by 1998 (2008 for developing for different countries at varying points incountries). At the time of the London time by different consultants cannot beAmendments, there was still great scope for directly compared. The different Countryoptimization of alternative strategies.Flowt aion t Cpernhagve stamedens, Programs reflect the development over time inFollowing the Copenhagen Amendments, --- sototosadcssa ela

however thsi.olnertecs o phaseout options and costs as well ascountries in transition who must meet the individual experts' different views as to thecountries 1996 phasoutransiti o mstemedatte most plausible and recommendable choice ofstrict 1996 phaseout date and intermediate alternative technology, which is often basedreduction requirement of 75 percent by 1994. on practices and experiences in neighboringAnalyses show that in these countries there are regions. The substitutes recommended forfew options but to implement phaseout actions Tunisia and Czechoslovakia reflect thein all user sectors as soon as technically and technical solutions that have been adopted in

institutionally feasible. The actual difference the European Community and especially in thein alternative phaseout scenarios has all but Nordic countries. As an example, wecompletely disappeared. recommend CO 2 or water blowing as a

Today, it would be more relevant to recomme f or wate ble amelaborate cash flow analyses for these substitute for CFC- 11 in the flexible foamelaboratriesth cash flow drlysesed for tsector, whereas others have recommendedcountries that can be directly used for methylene chloride.sequencing investments and estimating levels It should be emphasized that Countryof reimbursement necessary from the GEF. ' Programs on ODS phaseout have already beenThis also means that project preparation completed for many countries. The next stepshould be given higher priority in the initial is project preparation, feasibility studies, andplanning phase. The question of heavy up- appraisal. Therefore, it is important that a costfront investment costs is critical in these catalogue focus on detailed project costs. This


catalogue can then be used as a guideline both policies, including strategies to phaseout theto calculating country incremental costs and as transitional HCFCs. The model or model typea basis for project preparation and feasibility can also be used to solve similar problems,studies when adjusted properly to accom- such as cost-effective reductions of green-modate local conditions and needs. A cost house gas emissions, S02 emissions, nutrientcatalogue should not be used for making desk pollution of water bodies, and many otherstudies, since local, supplementary data environmental problems.collection and industry contact are very On the other hand, the model is not wellimportant parts of the Country Program. suited for sequencing investments or forOften, the collection of cost data is the first estimating levels of reimbursement necessarycoordinated and comprehensive action taken from the Multilateral Fund or the GEF. Forto mobilize, inform, and motivate the local these purposes, a cash flow model developeduser industries on matters related to the in spreadsheet form would be more useful; theMontreal Protocol. two could easily supplement each other.

Another area where more research isneeded is the projection of future ODS prices Making Calculations More Transparentas well as prices of substitute chemicals. This

is~~~~~ ~ th ke.atrdtrinn h piu The model calculations are, admittedly, notvery transparent. More transparency can be

phaseout strategy in developing countries, obtained by better explaining the parameters,Although this area is extremely difficult to say variables, and equations in the model, bymuch about, the basis for guessing could beimproved, for example, through contacts to all organizing the model in a morethe major ODS producers in the World (to find udrtnal a,adb lmntnoth wato thei production pln ared wto calculations that are not used later on.

However, it should also be stressed that thethe supply costs of altemative chemicals are). model (regardless of whether or not it is made

Although priority should be given toAlthough priority should be given to more transparent) is not suitable for use by

eliminating large ODS quantities as soon as . .possble mor reearh isor illbe nede in persons unfamiliar with the General Algebraic

possible, more research is or will be needed in Modelling System (GAMS) and economicthe many small applications of ODS- modeling and linear programming techniques.especially MCF, CTC, and CFC solventsapplications and the options and costs for Bibliographyreplacing them. Finding alternatives to methylbromide has also now become a new priorityfor which research and data gathering is and the World Bank. 1993. "Regionalneeded. n h ol ak 93 Rgoa

Considerations for the Phasing Out ofOzone-Depleting Substances in Central and

Usefulness of the Model in Relation to the O e tgEastern Europe." Draft of Final Report,February. Prepared by cowiconsult as part

Basically, the CFC cost model is an of the Environmental Action Programmeextremely powerful tool when it comes to for Central and Eastern Europe.testing alternative ODS phaseout strategies and


Table 5-l. List of ODS References

Project descriptiotn Country / year / clienzt

Country program for ODS phaseout. Russian Federation.1993 - ongoing.Ministry of Environment Russia; DanishMinistry of Environment .

Regional considerations for the ODS phaseout. Central and Eastern Europe.1992 - ongoing.The EC: the World Bank.

ODS assessment and country strategy. Bulgaria.1992 - ongoing.Ministry of Environment Bulgaria; the EC.

ODS assessment and country strategy. Hungary.1992 - ongoing.Ministry of Environment Hungary; the EC.

Country program for ODS phaseout. Czechoslovakia.1992.Federal Committee for the Environment,CSFR; Danish Ministry of Environment; theWorld Bank.

Country program for ODS phaseout. Tunisia.1992.Ministry of Environment, Tunisia; theWorld Bank.

Country program for ODS phaseout. Ghana.1992.Ministry of Environment, Ghana; UNEP.

Feasibility study of the auto air conditioning CFC recycling Thailand.Program. 1992 - ongoing.

Ministry of Environment Thailand; theNWorld Bank.

Study of substances that deplete the ozone layer. Philippines.1990-1991.Ministry of Environment of the Philippines;the EC.

Recovery of halons for destruction and recycling. Denmark.1991.The Environmental Protection Agency,Denmark.

Consumption and application of CFCs and Halons. Denmark.1987-1992 - ongoing.The Environmental Protection Agency,Denmark.

Information project on alternatives to CFC- 113 and methyl Denmark.chloroform in the metal degreasing industry. 1991.

The Environmental Protection Agency,Denmark.

Assessment of matters related to the revision of the Montreal European Community.Protocol on substances that dteplete the ozone layer. 1989-1990.

The European Commission (DGCxI).


Project description Country / year/ client

Alternatives to CFC- 113 in the dry cleaning sector. The Nordic countries.1992.The Nordic Council of Ministers.

Technical and economic consequences of reducing CFCs. The Nordic countries.1986-1988.The Council of Nordic IndustrialFederations and the Nordic Council ofMinisters.

Trade aspects of regulating CFC use. The Nordic countries.1988.The Nordic Council of Ministers.


Table 5-2: Summary of Growth and Other Assumptions Madefor Tunisia

Unconstrained CFC use199)

Area/user sector Applicaticn (metric tons) Growth and other assumptions Period

Population 8.2 million' 19912.3% average annual growth 2 1991-19972.1% average annual growth 1997-2010

Households 1.6 million 1991

Real GDP 5% average annual growth3) 1991-19973% average annual growth 1997-2010GDP/capita: USS 1,560 1991

Refrigeration and air Refrigerators for the 20 10% average annual growth 1991-1997conditioning local market 5% average annual growth 1997-2010

Domestic refrigeration Refrigerators for 40 5% average annual growth 1991-1997export 3% average annual growth 1997-2010

Servicing (average 125 10% fail 1' yearrefrigerator failure) 5% fail 2n-9'th year

90% fail I 0" year10% fail I ph year8% fail 12'-20h yr.Average lifetime 20 years

Production of 50 5% average annual growth 1991-1997refrigeration 3% average annual growth 1997-2010equipment, Average lifetime (comm.) 7 yearsinstallation and Average lifetime (ind.) 20 yearsservicing

Commercial and Servicing < 5 3% average annual growth 1991-1997industrial refrigeration 3% average annual growth 1997-2010and air conditioning Average lifetime 10 years

Mobile air conditioning

FoamRigid PU insulation in 60 10% average annual growth 1991-1997

domestic 5% average annual growth 1997-2010refrigerators for localmarket

ru insulation in 100 5% average annual growth 1991-1997domestic 3% average annual growth 1997-2010refrigerators forexporn

Other Pu foam and 25 5% average annual growth 1991-1997packaging foam 3% average annual growth 1997-2010

Flexible Mattresses and 400 7% average annual growth 1991-2010cushions

Aerosols' All applications 300 4% average annual growth 199 1-19975% average annual growth 1997-2010

Solvents Various minor < 5 5% average annual growth 1991-1997applications 3% average annual growth 1997-2010

TOTAL CFCs 1,130

Siz,rc e: Institut National de Statistique.' Source: The World Bank.'Source: The World Bank.The average lifetime of a compressor is 7-10 years, and the price of a new refrigerator is US$500.Unconstrained consumption of CFCs in aerosols differs from the actual consumption in 1991 because parl of the aerosol industry has already shifted

to altemrauve propellants. such as liquified petroleum gas (LPG).


Table 5-3: Assumed Substitutesfor CFCs in Tunisia

Assumed Unitearliest start substitution

date in costUser sector Current use Substitutes assumed Tunisia (US$/kg)

Domestic refrigeration cFc-12 HFc-134a 1995 40(production)Servicing of domestic CFC-]2 Nitrogen and recycling at 1993 1refrigerators large service shops

HFc--134a 1995 4Commercial refrigeration cFc-12 Better maintenance and 1993 1.5

CFC-1 15 (in service proceduresR-502) HCFC-22 1993 2

HFC-134a 1995 14Industrial refrigeration CFC- 12 Better maintenance and 1993 5

service proceduresAmmonia 1993 0HCFC-22 1993 0HFc- 134a 1995 5

Mobile air conditioning CFC-12 Better maintenance and 1994 2service proceduresHFC- 134a 1994 10

Polyurethane (Pu) foam CFC-I 1 50% CO2/water 1993 1domestic refrigerators 50% HcFC-142b/22 1995 3Other Pu insulation foams CFC-l 1 100% CO2/water 1993 1Polystyrene packaging CFC-12 Product substitute 1993 0

HcFc-142b/22 1995 3Flexible foam cFc- I1 CO,/water 1993 0.5Aerosols CFC- I 1/1 2 LPG and other alternative 1991 0

propellantsSolvents CFC-1 13 Aqueous cleaning 1995 4

MCFCTC

Fire extinguishants Halon-1301 INERGEN, FE-25 or other 1997 3alternative


Table 5-4: Costs and Benefits of Alternative Phaseout Scenarios in Tunisia

ODS useeliminatedover theperiod Early Total Discount

1991-2010 Incremental Incremental replacemen incremental ed costODP tons operating capital or t costs costs perkg(% of total costs project costs (million (million ODS

unconstrai (million US$ (million US$ US$ 1991 US$ 1991 eliminatened 1991 present 1991 present present present d

Scenario demand*) value) value) value) value) (US$/kg)I. Allowable 22,500

phaseout (54%) 15-20 1 95-100 110-120 5.12. Minimum

cost 23,200phaseout (56%) 20-25 2 1 20-25 1.1

3. Accelerated 35,000phaseout (85%) 25-30 2 1 25-30 0.8

*Total unconstrained ODS demand over the period 1991-2010 is estimated at 41,300 ozone-depleting potential (ODP) tons.

Table 5-5: Capital, Operating, and Forced Early Replacement Costs for Each Scenario for Tunisia

Scenario I Scenario 2 Scenario 3Allowable Minimum cost Acceleratedphaseout phaseout phaseout

(million US$) (million US$) (million US$)Increment capital or project costs 1 2 2

Incremental operating costs 15 22 26Forced early replacement costs 98 1 1for refrigeratorsToTAL 114 25 29

Table 5-6: Summary of Recommended Government Actions in Tunisia

Sectior Intended Estimanted costsYear(s) Description o)faction (if any) eftct/purpose USS 19921993-1995 1. Creation of an Ozone Committee for All Prerequisite for effective 210,000

implementation of the Action Plan implementation of ODS phaseout inTunisia

1993 2. Monitoring system for ODS use by All Monitoring and control and basis for 60,000substance and end user sector efficient and considerate policy

action1993 3. Introduction of regulatory measures All Control, incentive, and revenue 60,000

generation1993-1995 4. Introduction of an accreditation system Refrigeration sector Incentive for conserving ODS during 60,000

for refrigeration technicians service of refrigeration equipment1993-1995 5. Information dissemination User industries and the Awareness, ODS use reduction and 30,000

general public conservationTOrAL 420.000

* The distribution of government action costs on individual items is arbitrary, since the Ozone Committee secretary willbe involved in all activities.

MI USD Metic tons

-T I I I~~~~~~~~~2I

I /~~~~~~~~~~~~~~~~Z

I .,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.

It~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Z

0~~~~~~~~~~~0

C~~~~~~~~~~~

I a 8aa IIsnii

The World BankHeadquarters European Office Tokyo Office U1818 H1 Street, N.W. 66, avenue d'l6na Kokusai BuildingWashington, D.C. 20433, U.S.A. 75116 Paris, France 1-1 Marunouchi 3-chome

Chiyoda-ku, Tokyo 100, JapanTelephone: (202) 477-1234 Telephone: (1) 40.69.30.00Facsimile: (202) 477-6391 Facsimile: (1) 40.69.30.6( Telephone: (3) 3214-5001Telex: MCI 64145 WORLDB3ANK Telex: 640651 Facsimile: (3) 3214-3657

MCI 248423 WORLDBANK Telex: 26838Cable Addiess: INlMAFlAU

WASH lN(; rs lNDC.

The Global Environment Facility (GEF)Secretariat

1818 E Street, N.W.Washington, D.C. 20433, U.S.A.

9 780821 331330

Cover design by Soleil Associates ISBN 0-8213-3133-7

1 4c(4 4 public disclosure authorized - world bank · 2016. 8. 5. · hcfc hydrochlorofluorocarbon...

Documents