United States Environmental Protection Agency Air and Energy Engineering Research Laboratory Research Triangle Park NC 27711 Research and Development EPA/600/S2-89/021 Aug. 1989 Project Summary Continuous Performance Monitoring Techniques for Hazardous Waste Incinerators Rachel K. Nihart, John C. Kramlich, Gary S. Samuelsen, and W. Randall Seeker The report gives results of a study to determine the feasibility of an in- cinerator performance measuring methodology based on real-time continuous exhaust measurements of combustion intermediates; i.e., carbon monoxide, total hydro- carbons, and methane. The key issue was the correlation that exists between destruction and removal efficiency (ORE) and these inter- mediates. The study consisted of five steps: 1. A review of methods for moni- toring intermediate species in the exhaust gases. 2. Selection of instruments for eval- uation. 3. Evaluation of the instruments for response and potential interfer- ences. 4. An experiment in which test organic compounds were incin- erated in a laboratory-scale turbu- lent diffusion spray flame. 5. An analysis of the exhaust gas for both destruction efficiency (DE) of the waste compounds and the emission of intermediate species over a range of operating condi- tions from high- to low-efficiency operation. This Project Summary was devel- oped by EPA's Air and Energy Engineering Research Laboratory, Re- search Triangle Park, NC, to announce key findings of the research project that is fully documented in a separate report of the same title (see Project Report ordering information at back). Introduction A number of programs sponsored by the U.S. Environmental Protection Agency (EPA) have shown that thermal destruction is an effective technique for eliminating organic hazardous waste. Both field tests of operating incinerators and bench scale tests on turbulent diffusion spray flames have indicated that properly operated incinerators are very efficient. They have been found to destroy compounds to typically > 99.99% destruction and removal effi- ciency (ORE) when operated correctly. However, there is currently no way to continuously monitor the ORE perform- ance of the hazardous waste incinerator. More than 300 organic wastes have been identified by the EPA as being hazardous. A particular waste stream may consist of a number of these com- pounds, and assurance must be given that the incinerator is not releasing any of them during normal operation. Because of the wide range of compounds and the low concentration levels that need to be measured, direct real-time continuous monitoring is beyond the state-of-the-art of measuring techniques. EPA's licensing procedure requires that high efficiency operation (ORE >99.99%) be demon- strated prior to operation for a selected set of constituents in the waste stream. After licensing, the incinerator is main- tained within the range of operating conditions stipulated in the permit. During ------- routine operation, it is desirable to monitor the destruction efficiency (DE) of the unit and thereby allow corrective action to be initiated at the onset of unsafe performance. Methods available to measure specific hazardous waste com- pounds that are used during licensing trial burns (e.g., organic trapping and gas chromatography separation and analysis) are suitable for routine performance monitoring. They are both skill- and labor-intensive, noncontinuous, and the results are not immediately available after sampling. Thus, there is a need for the devel- opment of an indirect real-time con- tinuous monitoring methodology for haz- ardous waste incinerator performance. A number of indirect methodologies have been proposed. For this discussion, the potential methodologies can be cate- gorized as: • Tracer measurements (within waste stream composition or added). • System parameter measurements (pressure, temperature, flow rate). • Combustion intermediate species measurements (e.g., CO, hydro- carbons). • Combustion product measurements (C02, 02, NOX). Methodologies which rely on these measurements are largely untested and not completely developed. This study will concentrate on the use of intermediate species as a measure of incinerator performance since they appear most feasible at this time. Field tested and relatively low cost instrumentation is commercially available for several of the stable species. Combustion of Organics The combustion of organic molecules is a complex series of elementary reaction steps involving a myriad of intermediate species. For example, to completely describe the oxidation of the simplest organic (methane) requires more than 100 reactions and involves 25 inter- mediate species. For more complex organics, the numbers increase geo- metrically with the molecular weight. However, only a few of the species are stable; e.g., lower molecular weight hydrocarbons, methane (CH4), carbon monoxide (CO), and formaldehyde. Also, these stable intermediates are common to the combustion of most organic compounds including hazardous waste and auxiliary fuels. Organics burn in a complex series of fundamental reaction steps that even- tually lead through CO as the principal intermediate before the completely oxi- dized state (C02) is reached. CO is a thermally stable compound, compared to hazardous waste organics. Since it is a intermediate, large amounts of CO are produced in the flame zone that will be eventually destroyed if mixing is good. If mixing is poor, CO can escape the flame by mechanisms of thermal quenching. Excessive CO emissions usually reflect poor mixing or incorrectly set combustion air. The CO level measured in the stack depends on many factors; e.g., waste properties, spray nozzle design and operation, combustion air mixing pattern, and incinerator chamber design. CO can be measured continuously, using nondis- persive infrared analyzers, to concen- trations of a few parts per million (ppm). Exhaust emissions of hydrocarbons have three sources: (1) the unburned hazardous organic constituents due to a low DE, (2) unburned hydrocarbons from the auxiliary fuel, and (3) intermediate hydrocarbons generated during the com- bustion of the hazardous waste and auxiliary fuel. Which of these is dominant depends on the operating condition of the incinerator. However, the dominant source is likely the intermediate and, in particular, the light intermediates such as methane that are typically more thermally stable than the hazardous and fuel constituents. Total hydrocarbons can be measured with a flame ionization detector (FID) down to a few ppm, while CH4 can be measured with similar sensitivity using a nondispersive infrared analyzer (NDIR). Formaldehyde, another hydrocarbon intermediate, is somewhat less stable than CO or CH4 and no suitable instru- mentation is currently available for routine continuous monitoring. Study Objective The objective of this study is to determine the feasibility of an incinerator performance monitoring methodology based on real-time continuous exhaust measurements of combustion intermedi- ates, specifically: CO, total hydrocarbons, and CH4. The key issue is the correlation between destruction and removal effi- ciency (ORE) and these intermediates. The study involved five steps: 1. A review of methods for monitoring intermediate species in exhaust gases. 2. Selection of instruments for evalua- tion. 3. An evaluation of the instrument response and potential interfere™ 4. An experiment in which a set of organic compounds are incineratt a laboratory-scale turbulent diffu spray flame. 5. Analysis of the exhaust gas for DE of the waste compounds and emission of intermediate species i a range of operating conditions I high- to low-efficiency operation. Study Approach This approach was designed determine if correlation exists betw the level of combustion intermediates the DE in the turbulent spray flame. turbulent flame reactor (TFR) can operated under conditions where D are very high (>99.999%). The 1 conditions investigated covered a rai of low-efficiency conditions and f failure modes: fuel rich, excessively i lean, poor atomization quality, and fla quench. These failure modes were c sidered to be representative of pi operation conditions that might exist liquid injection incinerators. The exhaust concentrations of the t hazardous waste compounds were me ured along with the concentrations of tc hydrocarbons, CH4, and CO for each the conditions delineated above in on to assess the correlation between DE a the concentration of intermediate specii The test compounds were mixed to <. by mass with diesel fuel and introduc to the reactor. The compounds usec acrylonitrile, benzene, chloroform, a chlorobenzene—were selected becau they are EPA-listed hazardous orgar compounds and because they are e pected to represent a broad range incinerability behavior. The purpose of this study was determine the feasibility of utilizing re. time continuous exhaust measuremen of combustion intermediates as a way monitor incinerator performance. The k< issue was to determine if a dire correlation exists between DE ar intermediate species concentration mea urements. DE was based on exhau measurements for specific input was compounds using a Tenax cartridge fi capture and gas chromatography, flair This report distinguishes between wastt destruction efficiency (DE) and waste destructior and removal efficiency (ORE). DE is based or measurement of input and output concentrations of a species across a thermal treatment reactoi but upstream of any control device which removes that species form the flow exhaust. ------- ionizatio.T detection for sample analysis. The intermediate species measurements were selected to be continuous and real- time. A comparison of the potential intermediate measurements with the availability of commercial instruments resulted in the selection of a Beckman 402 Total Hydrocarbon Analyzer (THC), a Beckman 864 NDIR methane analyzer, and an Anarad AR-500 NDIR CO analyzer. The response of each instru- ment was compared to DE measure- ments from a turbulent spray flame in order to experimentally determine if flame zone correlation exists. The TFR was designed so that the processes controlling DE in the flame (e.g , atomization quality, ballistic trans- port, turbulent mixing, flame quench, and flame-surface impingement) simulate those that control DE in the flame zone of a liquid injection incinerator The cold walls of the TFR were designed to emphasize flame zone performance by quenching post-flame reactions. As such, the TFR data do not directly address how post-flame processes such as after- burners and gas cleaning equipment affect the flame-zone correlations. The data presented are cross-plotted to produce the DE/continuous monitoring correlations shown in Figure 1 These represent a broad range of operating conditions form high flame efficiency to low DE failure conditions. Two general correlations were observed between DE and the intermediate species measure- ments: 1. THC and CH4: The correlation be- tween DE, THC, and CH4 is nearly proportional; i.e., a linear increase in the intermediate concentration cor- responds to a proportional increase in the hazardous component concentra- tion. 2. CO: The CO correlation indicates that a significant increase in CO emissions is necessary before the exhaust concentration of waste com- pounds increases substantially. Fundamental combustion kinetic studies (e.g., laminar flat-flames) indicate that hydrocarbon flames can be divided into two partially overlapping regions. In the first,, hydrocarbon fuel is rapidly con- sumed by reaction with flame radicals (0,H,OH) to produce CO and water. In the second region, the CO is oxidized to CO2 at a relatively slow rate. The flame can be made to operate inefficiently by, for example, a reduction in temperature or the addition of flame inhibiting compounds. The first manifestation of this inefficiency is tho release of CO because the hydrocarbon destruction reactions remain sufficiently fast to quantitatively remove organic molecules. It is only after the flame has become extremely inefficient that hydrocarbons are released, by which time the CO emissions are substantial. Although £ direct comparison between laminar flat-flame and turbulent flame results is difficult due to the fundamental difference between the processes, the reactions occurring in each are the same. These reactions are simply super- imposed on different fluid dynamic back- grounds. Thus, a qualitative explanation for the flame zone correlations observed here would include: • Since the hydrocarbon intermediates and the model waste compounds are made up of organic molecules, the flame destruction rate of each is similar, at least relative to the slower CO destruction rate. • Decreased flame efficiency is evi- denced by CO release while hydro- carbon intermediates and the waste compounds are still quantitatively destroyed. • Further decreases in flame efficiency bring about a concurrent release of hydrocarbon intermediate and organic wastes due to the (relative to CO) similarity in their flame destruction rates. Conclusions These data support the following conclusions: • A turbulent spray flame operating without afterburners or postflame gas cleaning can achieve a DE of 99.99% This implies that careful design of efficient burners can cause the flame zone of a liquid injection incinerator to perform most, if not all, of the legally required waste destruction. • The data indicate that flame conditions which minimize CO, THC, and CH4 emissions result in optimum waste DE. This indicates that maximum flame efficiency, defined by the release of fuel and fuel fragments, also results in maximum waste destruction. • Less than optimum DE performance was accompanied by an increased release of intermediate species. This means that, for all conditions examined in this study, a change in DE toward lower efficiency was always accompanied by an increase in CO, THC, and CH4. • The range of conditions for optimum flame performance, as defined by CO, THC, and CH4, was found to be less than or equal to the range for high DE. This means that, under some condi- tions, the flame performance could decrease withoc a significant de- crease in DE. • CO was found to increase signiricantly under some conditions in which the DE remained high. Application of these results to the continuous monitoring behavior of a full- scale incinerator requires seme assess- ment of how the postflame zone process, such as afterburners or scrubbers, affect the flame zon3 correlations. Due to the very low solubility of hydrocarbons and CO in aqueous solutions, incinerator scrubbing systems would not be expected to alter the continuous moni- toring correlations for organic wastes. However, the afterburner may be expected to impact the postf'ame zone correlations. The aliphatic hydrocarbons that make up the bulk of THC, in particular CH4, are generally equally or more resistant to nonflame thermal destruction than the waste corr pounds upon which thermal testing has been performed. This indicates that monitoring THC and CH4 approach is potentially more conservative than indicate* by the flame measurements. Thermal testing has not been performed on CO. However, detailed chemical kinetic predictions indi- cate that the nonflame thermal destruc- tion rate of CO is approximately II .e same as that for hydrocarbons (as opposed to the flame destruction rates in which the hydrocarbons predomine e). Thus, no large alteration of the f! tie zone corre- lation would be expected. The correlations shown in Figure 1 suggest that CO is possibly an overly conservative indicator of flame perform- ance. That is, a,i incinerator shutdown procedure based on response of a CO monitor may be environmentally safe but economically impractica1 The results of this study suggest the following approach to incinerator monitoring and control: • Use CO as an indicator of flame per- formance, but not as an incinerator shutdown criterion. This should pro- vide a way to tune the flame operating parameters. • Use THC as a shutdown alarm to indicate potential waste compound release. Use of these two instruments in tandem provides, through CO, a way to tune for flame efficiency and, through THC, a way ------- to indicate incipient waste emission and direct shutdown. Quality Assurance/Quality Control requirements are applicable to this project. The data contained in this report are NOT supported by QA/QC docu- mentation as required by the U.S. Environmental Protection Agency's Quality Assurance Policy. 99.0 Excess Air, 3.8 Iph O Excess Air, 5.7 Iph Excess Air, 2.8 Iph Poor Atomization Quench Coils 1000 3000 5000 10,000 CO (ppm) 15,000 Excess Air, 3.8 Iph . O Excess Air, 5.7 Iph D Excess Air, 2.8 Iph Poor Atomization O Quench Coils 99.0 15,000 Excess Air, 3.8 Iph O Excess Air, 5.7 Iph D Excess Air, 2.8 Iph Poor Atomization Quench Coils o.oov 99.5 99.9 100.0 99.0 150 250 350 450 550 CH4 (ppm) 650 99.5 99.9 100.0 U. U. W Figure 1. Fraction of hazardous compound remaining m the exhaust as a function ol intermediate species concentrations. ------- R. Nihart, J. Kramtich, G. Samuelsen, and W. Seeker are with Energy and Environ- mental Research Corp., Irvine, CA 92714-4190. W. Steven Lanier is the EPA Project Officer (see below). The complete report, entitled "Continuous Performance Monitoring Techniques for Hazardous Waste Incinerators," (Order No. PB 89-195 192IAS; Cost: $15.95, subject to change) will be available only from: National Technical Information Service 5285 Port Royal Road Springfield. VA22161 Telephone: 703-487-4650 The EPA Project Officer can be contacted at: Air and Energy Engineering Research Laboratory U.S. Environmental Protection Agency Research Triangle Park, NC27711 United States Environmental Protection Agency Center for Environmental Research Information Cincinnati OH 45268 Official Business Penalty for Private Use $300 EPA/600/S2-89/021 CHICAGO ------- |