_____ ____20760 OLx,x3TION MANAGBEMrp i\A ° In Focu's DISCUSSION NOTE NUMBER 7 DECEMBER 1 999 Environm ental Among the potential effects of fossil fuel combustion are urban air pollu- tion, acid rain, and changes in global climate. A study of six cities in Challenges of developing countries and of an industrial area in Poland examined how Fuel Use various fuels and sources contribute to health and other environmental damages and analyzed abatement strategies for reducing these damages. Substantial differences were found between cost-effective strategies for mitigating local and global environmental effects. The study illustrates the difficulty of devising efficient policies, given the complex relationships among pollution sources, control options, and environmental impacts. This Kseniya Lvovsky real-world complexity calls for a skillful mix of policy instruments built on and Gordon Hughes rigorous analysis. Worldwide, exposure to the high levels cities around the world that suffer from US$3.8 billion, of which health impacts of particulates in urban air causes high levels of air pollution (see Box 1). accounted for 68 percent.1 Climate hundreds of thousands of cases of pre- The cities differ in geographic and change impacts amounted to 21 per- mature death and respiratory illness. climate conditions, demographic char- cent (using a shadow price of US$20 The levels of exposure and the associ- acteristics, fuel mix and use patterns, per ton of carbon emissions), and lo- ated health burdens are much higher the sectoral composition of the eco- cal nonhealth effects contributed 11 in low- and middle-income countries nomy, and income level. Together, they percent. than in rich countries. These country- have a total population of nearly 50 The key factors that influence the specific problems interact with a million and represent a span of vari- magnitude of environmental damages growing concern about global climate ables that affect the environmental and the relative shares of local and glo- change, which has no boundaries. De- costs of fuel use. The evidence emerg- bal impacts are the fuel mix and the signing policies and measures to ing from this exercise is likely to be sectoral composition of fuel use. Fig- combat the adverse environmental ef- representative of the typical situation ures 1 and 2 illustrate three important fects of fossil fuels is becoming an in many urban areas in developing findings. urgent challenge. countries. I . By far the greatest part of the local damages, which are dominated by In addressing this challenge, it is health impacts, comes from small essential to take account of the magni- Damages FueLs. household, commercial, and indus- tude of the damages attributable to and 5ource5 trial boilers and stoves and from different fuels, sectors, and pollutants. In the six urban areas, the social costs vehicles, rather than from large in- This note draws on studies of six large of all environmental impacts totaled dustries and power plants. Local RiAA29 Man,qen AWfs Notes tPMDNs) are part oi the knowtede mangrement effort of the Environment Family at the World anhk Grp. They are desiged to foder proessimol disnion, diseminte lessons leared from Bank operations, and trander beat practices in pollution magement. The views herein are those of tie authors) and are mt ofial Oky of, or attributable , the World Bank Group. The 1'tDN is an Open forum, Comments and suggestions for future publications should be directed to Maa Low nmloeioworldbank.org, or to room MC 5-1Z5. AuhOr lsniya Lvis an ernmntaL eDnomit in the World Bank's South Asia Environme Unit. Gordon Hughes s a Senir Advisor, Environment Department. Tis note is based on lseniya Lvovsky, Gdon Huges, David Madison, Dart Ostro and Dad Peae, Ah- FJi¢o sd the SodaO Ca f&e World Bank Technial Paper. forhcoming. Managing Editr; Magda Lovei, Environmnt Eonomist of the World Bak's Environment Department lENV<; Designer and Puliser: Jim CantrelL JENV7. THE WORLD BANK Box X If. XCoverage and Mthodologyof theSix-City Study fuels impose the highest social costs per B ~~~~~~~~~~~~~~~~~~ton of fueli2 The environmental costs Envronmental d a of fuse considered of these fuel uses are so high that the Adjverse heath 0efects of exposureto air polluItion (tfor example, inreased marginal damages exceed both pro respiratory illness andiprature death in the exposed population) 0LocalSnonheoltheffects (reduced Svisiblity8,dsoiling,and material damage)f: ducer and retail fucl priccs. Thc * Contribution ito,C glba climate chage greatest disparities between local and Pollutants 2 ti : :t 0t :$ ;t ;t ) 0i St fi :00i 0$0X000EtfC i0ji0i0i Q0 :global damages are also found for these • Locoal ects:. particulate mnatter (PM,0particles less than I10 icrons in aero- fuel uses. dynamic diameter), sulfur dioxide r(SO),and"niren66 oxides (NO) a Global effects: carbon dioxide M(02) Mitigation 5trategie5 Fuels Sim cply calculating damages is not an . Major fossil fuels coal fuel ol, diesel fuel, and gasol'ine) and wood adequate foundation for policv. The Sources of pollution next step is to compare the environ- Power plants, district heatinrgklplantus,lae indstrial and commercial boilers, small mental damages with the costs of the industrial andd commerciall boilers, houfsehold 'stoves'and boilers,0 and vehicles0 available control options. An analysis Cities of this kind was conducted for the Bangkok (Thailand), Kr (Poland) Manila (Phiilippines),Mumbai (India), Santiago metropolitan area of Katowvice in south- (Cfhile), and Shanghai (China) en Poland. The Katowvice area has MaFer seps inndamge ssessment historically relied on coal mining and each sectori and the quality Off the fuel(ash andsulfupr ontent of coal; isufur heaNy industrrv, and today it represents ecnten of petroleum productl-,sulf) a mnajor obstacle to Poland's compli- . Emissions estimates for all Ifuheluses bNasedon the fuel use invelntory and the ance with current and proposed abatement technologies, (eif any) adopedineach sctomr European Union (EU) directives for * Source apportionment that relates soure-s pecifc emissioins to the impact on ambient air quality ambient conditions and e el sign eling (including A vide range of control options simulation of secondary particlesrfromeSofandntOroemissions *a Assessment of health i pacts using dose-response relationships between am- across various pollution sources was bient levels of certain pollutants and health effects, as established in a large examined. The options included instal number of studies lation or upgrading of filters and dust V Valuation of mortality, morbidity, and nonhealth efcts using a coherent set of controls- improved combustion tech- estimates based on the willingness-to-pay approach, which provides a basis for comnparisonfacross differntenst effctsandcountriesi0000000:X0:000Q00\0;0000 niques; a switch from coal to alter- native sources of heat such as gas, smokeless coal, or district heating; and pollutants emitted by small (low- uses-and thus need to be designed fuel-saving measures that could be stack) sources are responsible for differently-according to whetlher the adopted in response to projected much higher marginal damage primary objective is to mitigate local or changes in the structure and level of costs per ton because the emissions global impacts. energy prices, assuming transition to are dispersed over a small area and very close to the exposed popula- As Figures 3 and 4 show. diesel- the EU pattern. tions. powered vehicles and small stoves and 'T'he analysis focused on three 2. By contrast, large sources, which boilers that burn coal or other "dirty" least-cost strategies: are the main contributors to C02 emissions, have a greater effect on Figure 1 - Sectoral contribution to local and global damages: Average global damages than do small for the six cities 1,200- sources. 3. The sectoral differentiation in fuel 1,000 * Global damage use is at least as significanit for the C 800 environmental costs of fuel com- - DI Local damage bustion as thc differences in the E 600 , type of fossil fuel used. (See the u 400 data for various uses of coal and for other fuels in Figure 2.) 0 r These results imply that policies 0 have to target different sectors and fuel Power plants Large boilers Vehicles Small furnaces Figure 2 - Fuel composition of local and global damages: Average for health damage to 2.5 million people the six cities in the area 1,200 _ - - - . - . *0 S2 an(d NOX eniissions-US$800 1,000 per ton and US$2,500 per ton, re- U Global damage spectivelv, based on estimates by the 800 .Local damage European Commission's ExterneE 2 600 1 project of damage to crops and for- to 400 i i E Eests caused by acid deposition and M 400 HzIe - ~~~~~ozone 200 J J J * * * * i * CO2 emissions-US$20 per ton of 200 K--= _ _ _ _ , _ _ _ __ _~ _carbon. Coal: Coal: Coal: Coal: total Fuel oil Diesel Gasoline Regardless of its objective, each power large small strategy yields a host of complemen- boilers boilers tary benefits from reducing local, regional, and global pollution. Figure 5 Figure 3 - Marginal environmental costs of fuels: Average usage in the shows the total social benefits that six cities would be generated by incurring 300 r------ - - different costs under these three strat- I t - l Costs of local pollution 250 | (average) egies. Costs of climate change For a particular cost level, the 200 benefits of the optimal strategy arc, by +spot market prices, 1 995 -* Spot market prices, 1995 ~~~~~definition, equal to or greater than the 150 benefits of any strategy that focuses on 100 one or two pollutants only. An interest- ing result is that the PM10 strategy is 50 t7 < - :-:very close to the Optimal strategy (and o4 .__ even identical at the lower part of the Coal Fuel oil Diesel Gasoline curves). This indicates that there is no practical difference between the two Figure 4 -Marginal environmental costs of coal across sectors: Krakow strategies until local pollution is sub- 160 stantially reduced. There are, however, 140 - Local costs considerable differences between the 120 _ Global costs .... ... P:NI Io0 strategy (or the Optimal strategy) 100 - . L and the CO2 strategy. At the beginning ,,so0 + Local price of the curves, there is a small ov erlap D 60 area because switching certain indus- 40 trial boilers from coal to gas is the most effective measure under both strate- 20 0 - _ . gies. The ranking of measures under Power plants Large boilers Small sources the strategies then distinctly diverges until the upper ends of the curves. The * A AW10 strategy for reducing aver- global) caused by all four pollut- use of smokeless coal or gas for resi- age exposuire to PM and the ants-PM10, SO2, NO,, and CO,. dential heating and of some dust associated health damage, wAithout unitThe damages (or benefits) per controls for district heating plants and unit of pollution (or pollution reduc- other industrial boilers are most cost- regard to any other benefits tion) were valued using the following effective for combating PMIO exposure. * A C02 strategy for reducing CO2 weights and pollution units: Under the CO2 strategy, fuel efficiency emissions, without regard to any * Ambient PM10 resulting from PMI0 measures at power and central heat- local benefits emissions and from the conversion of some SOanNOtseodr inplnsana vic fo calo * An "optimal" strategy, defined as o2 and NOx to secondary particulates in the atmosphere- gas at district heating plants, are pre- the most efficient combination of US$11 million per pg/m3 (Micro- ferred. measures for reducing the total so- gram per cubic meter) change in The cost implications of this dif- cial damage (local, regional, and annual average exposure, based on ference are significant. Efficient programs of measures to meet the Figure 5- Costs and total benefits of alternative strategies proposed EU air quality standards for c 800 2005 and 2010-30 and 20 ,um3 of 700 _ PM10 respectively- would cost less 600' than US$40 million and US$65 mil- ' 500 r lion per year under the PMIo strategy. 400 C 'T he costs of meeting the same targets e would be much greater-US$120 ! 300] , million and US$165 million, respec- ° 200 tivelv-if a strategy focusing on CO2 100 - emissions were adopted. For annual < 0 costs between US$40 million and 0 50 100 150 200 250 300 350 US$170 million, social benefits gen- Annual cost in US$ Million erated under the CO2 strategy are PM strategy CO strategy --- Optimal strategy more then US$100 million lower each - vear. This loss in social welfare repre- Note: The total benefit includes the benefits of reducing average exposure to PM 10 and emissions of SO NO sents the costs of missed opportunities 2' x' and CO2. associated with the choice of the CO2 strategy, even when the total social suring a substantial reduction will he standards, land-use regulations, and benefits of CO2 reduction measures large. For example, the incremental other command- and-control inter- exceed the costs. costofmeetingEU targetsforimprov- ventions, which are easier to design ing local air quality in a manner that in a differentiated manner, will con- baLancing LocaL and maximizes the reduction in CO2 im- tinue to play a major role in air quality ( LobaL Objectiv\es plies a value of carbon of about US$80 management. When the greatest part There is almost no differencebetween per ton. of (health) damages comes from ihe PM10 and CO2 strategies at the S small, dispersed sources that burn the an cO stratgiea t the From -5trategie.5 coal or other dirty fuels, provision of pointuat which the marginal benefrthe i- a oLicies viable incentives for cleaner fuels and timal strategy (shown by the vertical In low- and middle-income countries cleaner combustion technologies dotted line in Figure 5). At this point, with high levels of urban air pollution, through innovative financing schemes all three strategies are very close and the most cfficient strategy for reduc- also seems a promising approach. The yield largely comparable local, re- ing overall environmental damages specific mix of policy instruments will gional, and global benefits. The most would be to focus on mitigating local ultimately depend on the prevalence cost-effective way to reach that point health impacts, at least in the short of particular fuels and sources in a is to follow the optimal strategy, which to medium term. The implementation given city, on distortions in fuel mar- is in large part identical to the PM10 of such a strategy, however, remains kets, and on the maturity of institu- strategy. Thus, by focusing first on a difficult challenge. tions. Serious analytical work is as controlling local pollution-which is The large range of environmen- important as ever in formulating responsible for by far the largest dam- tal damages for different combi- policy advice for developing country age in developing country cities-and nations of fuels and sources, as well cities that are choking in smoky air gradually introducing stricter regula- as the large disparity between the while becoming involved in the glo- tions and more expensive abatement Optimal and CO2 abatement strate- bal agenda. measures, it would be possible to gies, shows that simple pricing meet, over the longer term, local, re- measures, such as a carbon tax, will gional, and global environmental be neither sufficient nor cost- 1. The assessment of damages is based objectives in a manner that maximizes effective. An efficient strategy for ad- on 1993 data. the overall social benefit. dressing local pollution problems 2. Diesel is treated as a motor fuel only Additional financial resources implies a high degree of differentia- Other uses of diesel are included in from the international community for tion in signals across sources and the fuel oil category. reduction of carbon emissions can ac- locations that gan,not feasibly be 3. European Commission, Directorate- celeratethereconciliationoflocaland achieved by fueI taxes or emissions General XII. 1995. ExternE global objectives. But the costs of en- charges alone. Emissions and fuel Externalities of Energy. Brussels. r .BmSaMbLa#G g.>.^ . ; E l.;M