ADVANCING WASTE TO ENERGY TECHNOLOGYDESIGN AND PERFORMANCE

OF EPI FLUIDIZED BED RDF-FIRED POWER PLANTS WORLDWIDE

Michael L. MurphyEnergy Products of Idaho4006 Industrial Avenue

Coeur d’Alene, Idaho 83815epi2@energyproducts.com

Energy Products of Idaho (EPI) has designed and installed numerous waste-to-energy systems in the UnitedStates and Europe, with others currently under development. Among the latest are a number of installationsin operation and/or construction in Italy. The design of the systems has advanced over time to stay ahead ofchanging regulations and performance requirements.

The paper reviews the evolution to-date of the EPI fluidized bed technology, specifically as it is adapted forrefuse derived fuel (RDF) fired waste-to-energy facilities. Basic design criteria, equipment configurationsand emission requirements over the past fifteen years are discussed. Operating data from the most recentfacilities along with design issues and performance improvements are presented. The recent facilities in Italyrepresent the most advanced design to meet the projected European Community standards for waste to energysystem emissions which are more stringent, in most cases, than comparable standards in the U.S.

Background

The potential for energy recovery from waste is well documented over the past three decades. Since the first“energy crisis” of the early 1970's, the energy potential from municipal waste has been studied and promoted,dignified and vilified, and, yet, even to this day, remains largely untapped. In spite of numerous waste-to-energy projects completed during the heyday of mass-burn technology of the 1980's, a substantial majorityof the world’s waste continues to end up buried in some multi-lined, leak-protected, surface-covered, off-gas-collected, hole in the ground, fondly known as a landfill. In a society marked by space travel, instantaneousworld-wide electronic communications, digital TV, neighborhood cell towers, and continued advances inevery form of technology imaginable, we somehow still maintain the illogical attitude that landfill disposalof wastes is the preferred solution.

Depending upon geography and socio-economy, the contents of a municipal waste stream can varysignificantly. In the U.S. and most industrialized nations where significant emphasis is placed upon “time,”the fast-food, microwave dinner, disposable mentality in the consumers goods industry has driven packagingand wrapping technology to new levels; thereby generating significant percentages of paper and plastic wastesin the disposal makeup. Combined with the putrescible wastes resulting from foodstuffs, vegetables, andother organic matter, the fraction of energy-containing materials in a typical waste stream can approach 75-90%. In less-developed regions, the putrescible fraction comprises the majority of the makeup, withpackaging materials much lower. A typical waste in the U.S. contains between ten and twelve million Britishthermal units (Btu) per ton, equivalent to nearly two barrels of oil. Considering the current U.S. populationin excess of 280 million, each generating about one ton of waste per year, this equates to over five percentof the annual consumption of oil in the entire country. To date, only about 15% of that energy is beingrecovered. Roughly speaking, there remains an as-yet untapped energy resource capable of generating 18,000

MW of electrical energy. Of course, not all of that is practical or economical to recover. The balance of thispaper discusses one technology currently capable of tapping into this vast energy resource worldwide.

Waste Control Options?

By its very definition, waste denotes a material with no further beneficial use. Hence, to minimize thequantity of waste generated today, much has been said and done to redefine the material and create new usesfor it. The motto of the solid waste industry over the recent decade has been to “reduce, reuse and recycle.”While numerous programs have been extremely effective in certain locations to reduce overall waste disposalrequirements, many of these programs have more effectively served only to “restrain and retard” long-termresolution of the problem. In many instances, recycling efforts have managed only to increase costs incollection and handling with much of the “recycled” material eventually ending up in the landfill due to alack of market.

Table 1 shows the typical composition of a municipal waste stream in the US. As can be seen, the wasterepresents a tremendous “source” of materials for reuse and recycling. With 280 million tons annually, ifonly 80% of the glass, iron, and aluminum is recycled, it is possible to recover nearly 35 million tonsannually. Not only do recovery and reuse of these materials represent savings in the raw materials theyreplace, but the processing costs to recycle these materials can be significantly reduced. For example, twentyaluminum cans can be recycled for the same energy consumption required to make one can from rawmaterials. While recycling is an option worthy of praise and promotion, it should not be considered thepanacea for all waste disposal. Considering the composition of Table 1, about 25 percent of the makeup,including ash, textiles and rubber/ leather, and food and yard waste, are unlikely candidates for recycling.Efforts to recycle rubber tires have been marginally successful in some instances, but account for a very smallportion of the tire quantities currently discarded. Recycling of paper and corrugated, plastics, and even somewood, has proven most successful; however, the recycling of these materials is still more of a delay tactic,putting off the inevitable disposal of the material until it has “cycled” through the loop a couple of times,ultimately breaking down or otherwise becoming less recyclable, until ending up in the “rejects” stream onceagain. Often, the cost incurred during the recycling process is substantially greater than the benefits gained.The fact that public support is strong even under such a negative economy indicates the degree of awarenessand desire to act for the long-term benefit of the environment, even when such actions may prove costly andmarginally effective. When looked upon from a life-cycle perspective, “recycling” may be, and probablyshould be, extended to include energy recovery as well. On a global basis, recycling the rubber and textileproducts, plastics, and much of the wood products will displace the consumption of considerable amountsof oil and natural gas in the overall energy equation. Because both oil and natural gas serve as the originalbuilding blocks for nearly all of the man-made plastics and rubber products we use today, energy recoveryfrom the product stream is essentially freeing up more of the “raw materials” of oil and gas for use in non-energy related applications.

Table 1Typical MSW Composition

(US)Component Percent

Ash/Grit 10.4Textiles 2.5

Paper/Corrugated 45.3Rubber/Leather 2.3

Wood 2.2Plastic 6.0

Food Waste 4.0Yard Waste 5.0

Glass 7.3Mixed Combustible 6.7

Ferrous 6.5Aluminum 1.8

Waste as Fuel

Given the above composition of typical waste, most of the items currently being recycled, namely glass, iron,and aluminum, are non-combustible and add no value to the energy content. Recycling these non-combustibles not only benefits those resources directly, but also benefits the remaining waste by theirabsence. The remaining mixture of corrugated, paper, plastics, and organic waste has increased energyconcentration. The energy content of the waste stream increases from a net of around 4,500 Btu/lb to nearly5,900 Btu/lb with the removal of the recyclable components and grit (dirt, sand, gravel, etc.).

Once the recyclable and non-combustible fractions are removed, the remaining blend is typically called refusederived fuel, or RDF. While not comparable to natural gas, oil or even coal in unit energy value, thequantities of material present a significant energy resource. From the 280 million tons of garbage generatedin the U.S. annually, approximately 75% or 210 million tons of RDF can be generated.

In terms of waste disposal, one of the most significant considerations is the volume required for landfill.Even though we often speak in terms of tonnages, in actuality, it is cubic-feet that is disposed of in ourlandfills. While densification, compaction, and other current practices have reduced the volumes associatedwith waste disposal, the impact of energy recovery on volume reduction is astounding. Because of the highdensity of the basic recyclable materials, once they are removed, the waste volume is nearly the same asbefore. After thermal oxidation, however, the waste stream is reduced to little more than ash, the residue ofthe noncombustible fraction of the stream. This typically is no more than 10-15% of the incoming weight.The density of the ash is four to five times greater than RDF, thereby yielding a reduced volume of only 2-3% of the original amount. Conservatively, recycling plus RDF thermal oxidation could extend existinglandfill life expectancies by 30 to 50-fold.

If it’s too good to be true, is it? If waste-to-energy is so good, and all the other disposal means are overratedand over-priced, why hasn’t it been utilized more? The answer to that not-so-rhetorical question is not asimple one. It may be the result of bad timing and bad press. Waste-to-energy, or at least waste combustion,

was initiated well back in the 1960's. At that time energy costs were just beginning to be economic factors.Waste disposal and landfill reductions were the prime motives. Unfortunately, at least for the RDFtechnology, some of these early plants experienced serious problems with their waste-processingtechnologies, some resulting in serious and sometimes fatal injuries. Such negative demonstration of thetechnology opened the way for mass-burn combustion wherein little or no pre-processing of the waste wascompleted. Nearly all incoming waste was introduced into the furnace and what came out was a mass of char,slag, glassed, waste which had little or no value and was transported to landfill. Because of the size, non-homogeneity and extreme variability of the feed stream, mass burn facilities were often over-designed andexpensive. While they proved to be very effective in waste reduction, they did not improve, tremendously,the perception from the earlier facilities that burning waste was difficult and costly. At that time thecompeting landfills were still considered fairly innocuous and were simple landfills with few environmentalrequirements.

With more recent knowledge of leachate, landfill gas generation, etc., the simplicity of landfilling has comeunder much greater scrutiny. Coupled with the public attitude toward greater recycling, environmentalaccountability, and rising uncertainty in the energy arena, conditions have become much more favorable forfull-service resource-recovery and waste-to-energy facilities. Because of the recycling initiatives, much ofthe waste stream is already subjected to processing, sorting, and removal of much of the non-combustiblecomponents as previously discussed. Advances in technology have greatly improved these processes andreduced the dangers experienced in the first generation. The remaining waste is almost fuel quality alreadyand can be converted to RDF with very little additional technology or cost involved. Once converted to afuel source, the RDF becomes much more suited to advanced thermal oxidation technologies such as fluidizedbed.

Another impedance to the widespread adaptation of waste to energy technology is the concern overenvironmental impacts. Much of the current database on plant emissions comes from mass-burn systems,many of which are early installations. Changes in requirements, technologies and even measuring procedureshave all contributed to current emission requirements which are much more stringent than levels achievedin many of the earlier facilities. With little or no demonstration of the required emission levels, establishingpublic credibility and support for these new facilities is often difficult at best. Fortunately, as moreexperience is gained on the newer waste-to-energy facilities and more research exposes the hazards ofexisting landfill emissions, the acceptance of newer plants, especially utilizing state-of-the-art thermaloxidation technologies, is increasing. Moreover, as more is known of the emissions and long-term hazardsof landfilling options, the technology and costs are also increasing. Where landfills present a perceivedsuitable alternative, their cost to site and develop are becoming more and more prohibitive. By includingenergy recovery in the waste management plans, the landfill may be forecast to last into the next centuryinstead of only the next decade.

State of the Art Thermal Oxidation

Among other items, the preceding section addresses how a waste management program can be enhancedthrough a combination of recycling and energy recovery. Few would likely argue the logic of such forvolume reduction; however, much uncertainty still exists surrounding the viability of the thermal oxidationtechnology. While discussion and philosophy may advance such discussions to some length, the final proofis in the performance. Fluidized bed thermal oxidation has emerged as a leading technology in demonstratingsuccess in all aspects of waste-to-energy technology. To date, Energy Products of Idaho (EPI) hasdemonstrated thermal oxidation, or co-combustion, of RDF in its fluidized bed systems for more than 15years.

Fluidized Bed Technology , or fluidization, is the term used to describe the characteristics created by passing

Figure 1

an air stream vertically upward through a bed of solid particles. The upward velocity creates a lifting orbuoyancy effect on the particles, resulting in the suspension of those particles within the air. As the airvelocities are increased above a minimum fluidization velocity, the particles are no longer held to normalsolid-to-solid contact and they can float and travel around with the air stream. The fluidized media exhibitsthe physical qualities of a fluid and looks like a pot of boiling water. The result is that the physicalcharacteristics of the mixture become very homogenous. Particle concentrations are uniformly distributedand temperatures are very consistent throughout the bed region.

This technology has been used in thermal oxidation of solid fuels for over thirty years. The turbulence of thesand creates an ideal combustion environment which results in continuous surface cleaning of the fuelparticle. Ash deposited from thermal oxidation as the fuel is burned is removed by the etching effect of thebed material. New surface is continuously being exposed to the combustion air, resulting in a much fasterand cleaner thermal conversion of the fuel particle.

In a typical EPI fluid bed thermal oxidation system, shown in Figure 1, the unit consists of a refractory-linedthermal oxidation chamber, an air distribution system, bed material and an above-bed thermal oxidation

chamber. The bed is comprised of an inert silica or alumina based material suitably sized to be fluidized bythe portion of the combustion air introduced as fluidizing air into the distribution system. The granular bedserves three major purposes:

1) It acts as a thermal storage medium to hold the heat within the system. In most instances,the system can be shut down for nearly 48 hours and be restarted with the residual heat fromthe bed material.

2) Once at temperature, it provides the ignition source for fuel that is fed into the unit.

3) It continuously cleans the surface of the fuel particle from ash buildup, exposing it tocombustion air.

The thermal storage enables the combustor to utilize much higher moisture content fuels because the bedprovides sufficient heat to evaporate the moisture from the fuel particle prior to thermal oxidation. WithRDF, this thermal stability balances the thermal oxidation and maintains uniformity both in temperatures andgas concentrations. Although some portions of the RDF, namely plastics, tend to ignite and burn veryrapidly, their effect is stabilized by the bed and offsets the slower evaporation and thermal oxidation of thewet paper and organic/vegetable matter also present. Ultimately this mixing eliminates any temperatureextremes within the combustor and ensures uniform thermal oxidation and optimum emissions.

Fluidized beds are renown for their ability to control acid gas emissions from sulfur. By using a sorbentmaterial such as limestone in the fluid bed, a chemical reaction between the calcium in the limestone and thesulfur from the fuel form a stable calcium sulfate (gypsum) product. This material becomes part of theparticulate stream and is removed from the system with the normal particulate gas cleanup devices.Consequently, the acid gas or sulfuric acid, that would result from the thermal oxidation of the refuse-derivedfuel can be controlled in a very cost-competitive method compared with other methods of acid gas scrubbing.As emission levels continue to be tightened, added acid gas scrubbing may still become necessary; however,the benefits of in-bed sulfur capture remain an advantage.

Typical operation of a fluidized bed thermal oxidation system requires an external air preheat system to heatthe bed sand up to a normal ignition temperature for the fuel. In most cases, this requires sand temperaturesof 800 to 1000°F. To minimize gas emissions during startup and support complete fuel thermal oxidation,additional gas burners are located in the furnace above the sand bed section to bring the vapor temperaturesup to 1600°F before introducing fuel. Once these temperatures are established in the furnace, fuel and airflows can be increased. As temperatures are stabilized the external fuel can be eliminated. This operatingphilosophy can accommodate a cold start-up in a matter of hours. Actual startup times are restricted byrefractory and boiler heat-up demands.

As fuel is fed into the combustor, Figure 2, a portion of its energy is released to the sand with the remainderabove the bed. Design of the combustor attempts to optimize the temperature profile through the system tohold the bed hot enough to ignite the fuel as it is fed and maintain vapor temperatures below the ash fusionlimits of the particular fuel. The primary temperature control comes from using excess air to dilute the outlettemperature to around 1700°F. Where fuel economy is desired, the amount of excess air is minimized. Underthese conditions, some heat release from thermal oxidation is removed directly from the fluidized bed regionthrough the use of in-bed steam generating surface. This heat transfer surface is placed in the fluid bed andis cooled with saturated water from the boiler circuit. As water circulates through these tubes, it extracts heatdirectly from the bed and creates steam in the water line. The heat removed from the bed results in loweroperating temperatures and allows the system to maintain temperature control with less air. For the samereason that the thermal oxidation of a fuel particle is extremely fast within the fluidized bed, the heat removalfrom the sand to in bed water tubes is also much higher than a normal boiler. Consequently, a considerableamount of heat can be removed from the system using a minimal amount of boiler surface. This allows foreconomical design as well as improved fuel efficiency.

Figure 2

In order for the fluidized bed to maintain its versatility under all fuels, it still must be operated using eitherchanging temperatures within the system or changing excess air levels to compensate for any variability ofthe fuel. This is typically not a problem under most operating conditions, but it does create designrequirements that must be considered when establishing fuel types, feed mechanisms, etc., to cover thespectrum of fuels planned for the particular plant. All conditions must be maintained such that under theworst possible fuel conditions the heat transfer surface within the bed will not remove more heat from thesand than is available from the fuel. The amount of in-bed heat transfer surface is dictated by the worst fueltype rather than the best. The worst fuel is typically defined as the fuel that has the highest moisture levelor the greatest volatile or fines content which results in most of the heat release being liberated above the bedrather than directly into the sand.

In a fluidized bed combustor, overall thermal oxidation efficiencies are extremely high, with carbon burnoutabove 99 percent.

Air EmissionsSince its first application in RDF thermal oxidation, the EPI fluidized bed has demonstrated its ability to meetthe required emission limits. Those limits have continued to tighten to the levels presented in Table 2 whichrepresent the basic emission requirements for Italy and much of Europe:

TABLE 2 Italian Emission Requirements (2001)

(Corrected to 11% O2)

mg/Nm3 ppm lb/M BtuCO (Carbon Monoxide) 50 43 1.35NOx (Oxides of Nitrogen) 200 170 0.22VOC (Volatile Organic Compounds) 10 5 0.011SO2 (Sulfur Dioxide) 100 38 0.11HCL (Hydrogen Chloride) 10 7 0.011PM-10 Particulate 10 0.005 gr/sdcf 0.011PCDD+PCDF (TEQ) .1 ng/Nm3

Products of incomplete combustion (PIC’s) are inherently lower from fluidized bed thermal oxidation systemsthan from the more conventional combustion technologies. When additional abatement of emissions isrequired, the versatility of the fluidized bed makes it easy to apply additional abatement technologies.Combustion at the proper temperature, excess oxygen and residence time limits CO emissions, but if thetemperature gets too high and there is too much oxygen, NOx production is increased. Time, temperature andturbulence are optimized within the fluidized bed combustor to achieve a balance between good combustionefficiencies and low emissions. Carbon conversion is maximized while maintaining the least possible excessair requirements within the system.

Carbon Monoxide (CO) emissions are also an indicator of the potential for dioxin and furan (PCDD &PCDF’s) formation. Dioxins, furans, CO, volatile organic compounds (VOCs) and polycyclic aromatichydrocarbons (PAHs) are all considered products of incomplete combustion (PICs). Emissions of all thesecompounds are significantly reduced in a well-designed fluidized bed furnace. These compounds aretypically the result of incomplete combustion caused by one or more of the following:

! Low and/or fluctuating temperatures ! Poor temperature distribution ! Wet fuel (on a grate system) quenching ! Inconsistent fuel feed ! Insufficient turbulence ! Insufficient residence time

For the most part PICs, particularly dioxins and furans are thought to be created in the combustion processdue to either poor furnace design or uncontrolled fluctuations in the fuel heating value and/or fuel feed rate.Such transient conditions can cause rapid devolatilization of waste materials, momentarily depleting orlowering local oxygen levels in the furnace. This can cause heavy transient loadings of unburned gaseous

and particulate matter. It is widely accepted that refractory- walled fluidized bed thermal oxidation systemsoperate at low levels of PICs including dioxins and furans due to their uniform and complete thermaloxidation characteristics.

Another key regulated emission is oxides of nitrogen (NOx). Formation of NOx in the combustion processcomes either from thermal fixation of nitrogen in the air at high combustion temperatures or throughconversion of chemically bound nitrogen contained in the fuel. Fluidized bed systems typically generatelower levels of NOx than other traditional combustion technologies burning the same fuels. The conversionrate of both “thermal” and “fuel” NOx is highly dependent upon temperature. Thermal NOx is generallyformed at temperatures above 2000/F in the presence of excess oxygen. Due to the inherent design of thefluidized bed system with its lower operating temperatures, typically below 1700/ F, “thermal NOx”generation is minimized. Operation at low oxygen levels can also help reduce the interaction of fuel-boundnitrogen with oxygen and resultant formation of NOx.

For further reduction of NOx, EPI has adapted selective non-catalytic reduction (SNCR) technologyspecifically for use in bubbling fluidized bed systems. SNCR technology works by adding ammonia (in theform of anhydrous ammonia, aqueous ammonia, or urea) into the vapor space of the thermal oxidationchamber within a specific temperature range. The ammonia (NH3) reacts with NOx to form nitrogen (N2) andwater (H2O). The chemistry is complex, but the basic equilibrium reaction is:

NO + NH3 ÿ N2 + H2O

The fluidized bed thermal oxidation system produces inherently low NOx levels. However, with some fuelscontaining high nitrogen levels, it is necessary to install SNCR technology to meet the strict emissionsstandards. The temperature profile in EPI’s fluidized bed system provides an ideal environment for successfulintegration of SNCR technology for the abatement of NOx emissions.

Abatement technology for reduction of SOx emissions is also easily integrated into fluidized bed systems.Sulfur present in the fuel combines with oxygen during the combustion process to form sulfurous/sulfuric acidwhen combined with condensed water droplets in the flue gas. The fluidized bed system uses a unique andvery effective process for reducing SOx. Sorbents such as lime, limestone, or dolomite containing calciumoxide (CaO) are added to the bed and react with sulfur compounds to form calcium sulfate or gypsum(CaSO4). This is a two-step process where SOx combined with CaO forms calcium sulfite (CaSO3) which isfurther oxidized to form gypsum. The reactions are as follows:

SOx + CaO+O2 ÿ CaSO4

Some fuels containing calcium have an inherent ability to abate sulfur due to the mixing action of thefluidized bed technology. This tendency can help to reduce or eliminate the need for additional abatement.

Control of hydrogen chloride (HCl) can not be accommodated within the thermal oxidation system as is thecase with sulfur dioxide due to the instability of the calcium chloride salt at the high furnace temperatures.The control mechanism is similar but must be completed at the lower gas temperatures at the exit of thesystem through the use of a spray dryer scrubber. Lime or lime slurry is introduced into the gas stream in areaction chamber where the acid gases remaining in the flue gas are reacted with the calcium to create an inertparticulate which can be removed from the system via a baghouse.

Mercury and heavy metals are neither created nor destroyed within the thermal oxidation process, but arecarried through either as solids or as vapor. Removal of these components is achieved most effectively byreducing the input of such contaminants in the feedstock. Current methods of waste processing have

Figure 3 - Northern States Power, French Island Power Plant

successfully removed a significant fraction as have battery recycling campaigns and other similar programs.Current abatement practices for removal of the remaining metals present in the outlet gases includeintroduction of activated carbon, especially for mercury control, and quench and scrubbing of the outlet fluegas stream to reduce temperatures below 200°F to condense any vaporized metals back to a solid particulatewhich can be removed from the gas stream.

As it is with all technology, the demands for bigger, faster, and better apply to emission control technologyas well. Where once the inherent control capability of the fluidized bed combustor plus some simple meansof particulate abatement were sufficient to achieve required emission control limits, the newer levels mandatestricter controls and additional measures to succeed. A brief review of the design parameters and operatingperformance of some of EPI’s RDF installations makes this very apparent.

OPERATING HISTORY - EXPERIENCE

Northern States Power - French Island Power PlantEPI commercialized its fluidized bed technology for RDF applications with the commissioning of the secondpower boiler conversion in the Northern States Power (NSP) French Island Station, Unit 1, in LaCrosse,Wisconsin, in 1988. At that time, simultaneous with the fluidized bed conversion of a 150,000 pph steamboiler, NSP incorporated an RDF processing facility to allow the facility to receive waste from the localcommunity to supplement the wood residue currently fired in the first boiler, also an EPI fluidized bedretrofit of another 150,000 pph power boiler. Upon completion, the total station generation capacity is 30MW of electricity using a 50/50 blend of wood and RDF, by weight. Steam conditions are 450 psig and 750F. Emission abatement and/or gas cleanup requirements included fluidized bed thermal oxidation for mostitems plus an electro-filter bed, also known as a gravel bed filter, for particulate.

The facility is shown in the accompanying picture, Figure 3. Results of various emission tests, published andunpublished, have documented the levels shown in Table 3 during the course of operation. These arecompared to the permitted limits, circa 1988.

Figure 4 - Tacoma, Washington Power Plant

Table 3French Island Power Plant

Emission Levels - Actual and Permitted(at 50% RDF co-firing)

Permit ActualCO ppm @ 10% O2 300 15-90TCDD TEF ng/dNm3 @ 7% O2 1400 0.68HCl uncontrolled ppm @ 8% O2 145HCl controlled ppm @ 8% O2 30 (proposed) 35NOx ppm (at 25% RDF blend) 120particulate (lb/ M Btu) 0.10 0.05

Tacoma Power PlantIn 1989, EPI retrofitted another pair of power boilers in the Tacoma, Washington, region to fire a combinationof wood, RDF, and coal. Steam conditions are 425 psig and 750 F. Permit requirements for this facilityincluded combustion control, in-bed limestone injection, and baghouse particulate control. This facility isshown in the Figure 4. The RDF portion of the fuel stream is typically less than 25% of the total butrepresents over half of the 600 daily tons generated by the Tacoma community. Source separation of

recyclables plus a refuse separation and fuel preparation facility provide the RDF used by the facility. Theretrofit of the original boiler facilities was achieved by converting each of the two, Sterling-type boilers toa heat recovery steam generator with a dedicated fluidized bed system located outside the boiler building.Hot exhaust gases from the fluidized bed plus steam generated from in-vessel heat exchange surfaces arerouted into the existing boilers to generate 265,000 pph of steam each. Each system is designed for an energyinput of approximately 415 M Btu/hr. Pollution control technology included limestone injection for SO2control in the fluid bed and a baghouse for particulate capture. The permitted levels for emissions are includedhere in Table 4 along with actual test results.

Figure 5 - Ravenna, Italy WTE Facility

Table 4Tacoma Power Plant

Emission Levels - Actual and Permitted

Permit ActualParticulate gr/sdcf 0.01 0.009SO2 lb/ M Btu 0.18 0.18NOx lb/ M Btu 0.50 ______CO ppm 425 1VOC lb/ M Btu 0.047 ______

Ravenna, Italy

A recent resurgence in RDF energy systems has seen the installation of one new facility in Ravenna, Italy,followed by another similar facility in Massafra, Italy. For the Ravenna facility, the system size is

approximately 7 MW electrical. Steam conditions of 600 psig and 715 F were selected. In this installation,all of the current European emission standards were required, plus provisions to meet the proposed standards.These emission levels follow those outlined in Table 2 above. To achieve these limits, the system includesthe following abatement controls:

1) Fluid bed thermal oxidation for CO, VOC, and other PIC’s2) SNCR utilizing aqueous ammonia for NOx abatement3) In-bed limestone injection for SO2 removal4) Multi-cyclone and baghouse for particulate control5) Spray dryer scrubber for SO2 and HCl

6) Activated carbon injection for mercury and dioxin7) Wet scrubber tower for final gas cleanup, heavy metals condensation and removal, andacid gas control.

The overall process design is shown schematically in the accompanying Figure 6.

The actual emission levels achieved during performance testing of this facility have been reported in otherpapers but are included in Table 3 herein for reference. As seen from the comparison of actual to permittedlimits, the performance of the system has far excelled in achieving the desired results.

Table 5Ravenna WTE Facility

Emission Levels - Actual and Permitted(all values in mg/Nm3 @ 11% O2 unless noted)

Actual Permit O2 9.9 >6%, volCO <1 50

Particulate <0.5 10SOx .1 100NOx 191 200HCl 2 10

HF + HBr <0.1 2VOC <1 10PAH 1 <0.003 0.1

PCDD + PCDF 0.034 0.1Total Heavy Metals <0.001 0.6

1 Concentrations in ng/Nm3

Massafra, Italy

Following on the success of the Ravenna facility, the Massafra power plant is nearly twice the capacity asthe earlier unit at 133,000 pph of steam at 667 psia and 752 F. The emission requirements are slightly morestringent for NOx (160 mg/Nm3), CO (40 mg/Nm3) and VOC (8 mg/Nm3), than was Ravenna; however, thecontrol technology for each remains more or less the same. This unit is currently under construction, withstartup scheduled for early 2003.

Figure 6Flow Schematic for Ravenna & Massafra, Italy