Analytical Techniques for MACT Industrial Boilers Emissions Standards
On September 13, 2004, the EPA promulgated Maximum Achievable Control Technology (MACT) emissions standards for industrial boilers. Emission limits were established for total selected metals (TSM), mercury (Hg), and hydrochloric acid (HCl). The heat content and moisture content of the fuel is also discussed in the rule. The regulations include site-specific fuel analysis plans to demonstrate compliance. The regulations specify approved procedures for sample collection, sample processing, sample preparation, and chemical analysis. The approved methods are shown in Table 6 of the boiler MACT rule. The regulations also allow the use of equivalent methods and alternative methods. In order to meet site-specific emission limits, equivalent and alternative methods are often used in the fuel analysis plan. When alternative test methods are used a written request seeking approval of its use must be submitted to the EPA. The compliance date for these industrial boilers is September 13, 2007. Compliance can be demonstrated by fuel testing or with stack testing in combination with fuel testing.
This article is divided into two parts: recommended solid fuel testing techniques and recommended stack testing techniques for compliance with boiler MACT regulations.
Columbia Analytical has developed analytical approaches to meet the industrial boilers emission standards. Analytical methods have been validated for solid fuels testing and stack testing. The recommended analytical techniques discussed in this article have met regulatory approval in many site-specific fuel analysis plans. Some of the rationale for method selection will also be discussed.
SOLID FUEL TECHNIQUES
Solid Fuel Collection and Processing Procedures
In order to obtain a representative sample, the regulations discuss both sample collection and sample processing techniques. There are procedures for collecting samples from a belt or screw feeder, from a fuel pile or truck, and from a falling stream. These sampling procedures describe the tools to take the sample, how the sample is taken, the amount sampled, and frequency of sampling. The collection techniques are listed in Section 63.7521(c) and Table 6 (ASTM D2234M-03, ASTM D6323-98, and any “equivalent” method). The processing techniques are discussed in Section 63.7521(d), a quartering technique. The result is one pound or more of a representative composite sample for subsequent laboratory analysis. The EPA mandates that a minimum of three composite samples are tested and three samples are typically submitted to the laboratory as part of the site-specific fuel analysis plan. More samples are submitted if a larger statistical population is desired.
Recommended Analytical Techniques to Comply with Boiler MACT Fuels Testing:
In order to perform the laboratory testing, each of the different steps used to process the samples needs to be considered to ensure that accurate and precise results are reported. These steps include:
• Taking a representative sample in the laboratory
• Particle size reduction
• Dry ash techniques
• Wet digestion techniques
• Instrumental techniques for TSM, Mercury, and Chloride
This discussion will focus on the recommended techniques used to reduce the particle size, to prepare the samples for instrumental analysis (dry ashing, wet digestion, oxidation), and to perform the instrumental analysis.
Taking a Representative Sample in the Laboratory
Typically a one or two pound sample of solid fuel is submitted to the laboratory for analysis. A representative aliquot of the sample is taken using the same quartering technique that is used in the field. The representative aliquot is then taken through some particle size reduction steps so the heat content, moisture, TSM, mercury, and chloride determinations can be made.
Particle Size Reduction Techniques
Samples must have their particle size reduced in order to take a smaller representative aliquot for each of the subsequent tests, which is the most important consideration when performing any of the chemical analyses. For TSM, NCASI has performed studies that demonstrated that the grinding equipment and screen size have an impact on the metals results. The samples should be ground fine enough to provide sufficient surface area to enable an efficient digestion. If the samples are ground too much, NCASI has shown that the grinding equipment can contaminate the samples with trace levels of chromium, manganese, and nickel (all TSM components).
Different approaches can be used to produce data that meets regulatory reporting limits without introducing metal contamination during the particle size reduction step of the process. We recommend two different approaches:
• Grind through a 6 mm and 1 mm screen after samples are air-dried at room temperature. The 6 mm screened sample is used in the dry ash technique for beryllium, cadmium, chromium, lead, manganese, and nickel. The 1 mm screened sample is used in the wet ash techniques for arsenic, selenium, mercury, and chloride.
• Cryogenic grinding of the sample using liquid nitrogen and agitation to reduce the particle size to a fine dust.
Digestion and Instrumental Techniques for TSM and Mercury
Because the industrial boiler emission standards are based upon the BTU content of the fuel, the desired method reporting limits to demonstrate compliance is variable. The lower the BTU content of the fuel, the lower the method reporting limit should be. With this in mind, we recommend methods and reporting limits for the worse case scenario. The two approaches shown in Table 1 meet the desired sensitivity requirements for low BTU fuels. Our Kelso, WA laboratory uses EPA Methods 6010, 7060, 7740, and 1631 to perform the testing. Our Jacksonville, FL laboratory uses EPA Methods 6020 and 7471. Both of these approaches along with the appropriate sample preparation techniques have been routinely accepted by regulatory agencies for Boiler MACT fuels testing. In both cases, the summation of TSM reporting limits is consistent (about 2 ppm) and meets sensitivity requirements for low BTU fuels.
Dried Sample (60° C) – Pre- 1mm Grind
Dried Sample (60° C) – Post 1mm Grind
All of the methods employed have a preparation step that dissolves the elements prior to the instrumental analysis step. The preparation steps used in conjunction with the instrumental method are shown in Table 2. Samples can be digested using a wet acid digestion technique or dry ashing technique in conjunction with the wet acid digestion technique. A dry ashing step is used to convert the elements to a form that improves their dissolution in the subsequent wet acid digestion step (EPA Method 3050B). Dry ashing is used for the six elements analyzed by EPA Method 6010.
Arsenic, mercury, and selenium are not performed using the dry ashing technique because of potential volatilization at the ashing temperature. There are specific acid digestion techniques to ensure that arsenic, mercury, and selenium are not lost in the analytical process. The wet ashing technique (EPA Method 3050B) is also used for the eight elements analyzed by EPA Method 6020.
Beryllium, cadmium, chromium, lead, manganese, and nickel are analyzed by inductively-coupled plasma (ICP) techniques in both approaches. The Kelso approach uses EPA Method 6010, an inductively-coupled plasma – atomic emissions spectroscopy (ICP-AES) method, after dry ashing of a larger sample. This method was chosen because it is cost effective (able to simultaneously determine multiple metals), obtains sufficient sensitivity for the MACT regulations, and is a well-documented method for reliability due to the larger initial sample aliquot and “clean” end matrix. The Jacksonville approach uses EPA Method 6020, an inductively-coupled plasma – mass spectrometry (ICP-MS) technique to obtain lower detection limits. This method was chosen because it is able to simultaneously determine multiple metals, is reliable, and meets the sensitivity requirements.
Arsenic and selenium can be analyzed by GFAA, ICP-AES, or ICP-MS techniques, but the dry ash step cannot be employed as it can for the other metals that are analyzed by these techniques. Generally, the ICP-AES technique is not sensitive enough for this application. That leaves GFAA and ICP-MS techniques. When analyzing these elements by ICP/MS, care must be taken to avoid a high bias. The primary issue is related to incomplete digestion of carbon-containing material (i.e. complete conversion of organics to carbon dioxide and water) resulting in a high bias during the ICP-MS analysis. If a significant amount of organic material is left behind as partially oxidized and/or hydrolyzed products, arsenic and selenium may be subject to increased ionization, which is manifested as a high bias.
Mercury’s critical level for low BTU fuels is around 0.036 ppm. The Kelso approach uses EPA Method 1631 (cold vapor atomic fluorescence spectroscopy) to achieve an MRL of 0.001 ppm. Jacksonville uses EPA Method 7471 (cold vapor atomic absorbance) with newer instrument technology that has been optimized to achieve an MRL of 0.005 ppm. Both approaches meet the sensitivity needs for compliance.
It should be noted that the sensitivity of each method is shown as a “Method Reporting Limit.” The Method Reporting Limit is demonstrated by successfully analyzing a standard at that level during the test and is further supported by current method detection limit studies. The method detection limit is a statistically determined limit usually performed annually under ideal laboratory conditions. When using data that is “non detect”, the method reporting limit should generally be used for interpreting the results, not the method detection limit.
If there are other applications or need for the data, there may also be a need for alternate approaches. For instance, in the NCASI grinding study, Columbia Analytical/Kelso employed ICP-MS techniques on the dry ash/EPA Method 3050B digestate to obtain lower detection limits to assess potential metal contamination by the grinding apparatus. Comparison of typical method reporting limits by the two techniques are shown in Table 3.
If there is a need for data on other elements (aluminum, iron, copper, zinc) that may be useful for operational purposes, they can also be obtained by EPA Method 6010 or EPA Method 6020 if specified upon submission of the samples. Instrumental methods that can simultaneously analyze several elements can be a very cost-effective approach to meet several program data requirements from a single representative sample. The ICP-AES and ICP-MS techniques can test for a wide variety of metals.
Digestion and Instrumental Techniques for Chloride
Chloride is analyzed from either the 1 mm screened material or the cryogenically ground material. EPA Method 5050 followed by EPA Method 9056 is used for the analysis of chloride. EPA Method 5050 is an oxygen bomb preparation method in which different chlorinated species are converted to chloride. EPA Method 9056 is an ion chromatography technique. Columbia Analytical has a 20 ppm method reporting limit and meets the critical value of 350 ppm for low BTU fuels. The ion chromatography method can test for a wide variety of anions (bromide, chloride, sulfate) if there is a need for these elements for other purposes.
Techniques for Higher Heating Value and Moisture Content
The higher heating value is determined from the 1 mm screened material or the shatter box prepared material. ASTM Method E 711-87 is an oxygen comb calorimetry method. Moisture content is determined on the sample before it is subject to particle size reduction by ASTM Method E 871-82. These analyses are straight forward if sufficient attention is used to obtain a representative sample.
These sample preparation and instrumental techniques have been routinely accepted by regulatory agencies for Boiler MACT fuels testing.
STACK TESTING TECHNIQUES
Recommended Analytical Techniques to Comply with Boiler MACT Regulations
If the MACT fuels testing results do not pass the fuels criteria, then source emission testing is performed for hydrochloric acid by EPA Method 26 and MACT multiple metals analyzed by EPA Method 29. In order to perform the laboratory testing, each of the different steps used to process the samples needs to be considered to ensure that accurate and precise results are reported. These steps include:
• Taking a representative sample in the field
• Documenting sample train and chemical
reagent cleanliness (field blanks)
• Segregating the sample train fractions
according to the method
• Different wet digestion techniques on the
sample train fractions
• Instrumental techniques on sample train fractions
• Calculations of reporting limits
This discussion will focus on the recommended techniques used for sample preparation and instrumental analysis of the source samples. Methods and reporting limits for MACT multiple metals used atColumbia Analytical Kelso are shown in Table 4. Please note that these are typical total micrograms (μg) reporting limits that are obtained for both the front half and the back half of the EPA Method 29 sampling train. The total microgram reporting limit and the amount of sample captured in each of the fractions are used to calculate the total emissions reporting limit.
Analysis for hydrogen chloride is performed by EPA Method 26. Samples are captured in an acidic impinger, neutralized, and analyzed by ion chromatography techniques. The total microgram reporting limit is dependent upon the volume of the impinger which varies dependent upon the moisture of the gas stream and amount of water used for rinsing. Typically impinger volumes are 50 to 100 mL. Reporting limits for chloride on typical impinger volumes are shown in Table 5.
MACT Metals Sample Preparation Techniques
Method 29 is very specific about how the different sampling train impingers and filters are handled, resulting in complex sample preparation steps. The sampling train consists of a quartz fiber filter and seven impingers. Samplers take the filter, impinger solutions, and rinseates from each sampling train to produce up to 15 containers for the laboratory to analyze (Containers 1, 2, 3, 4, 5A, 5B, 5C, 6, 7, 8A, 8B, 9, 10, 11, and 12). The laboratory receives these containers and processes them in 7 analytical samples (Analytical Fractions 1A, 1B, 2A, 2B, 3A, 3B, and 3C). The processes specified in EPA Method 29 for the different sampling impingers, sampling containers, and analytical fractions are complex, specific, and can be problematic for even experienced analysts performing the tests. This applies to both the field and laboratory staff.
The front half of the probe contains the filter media and is digested with a hydrofluoric acid technique to dissolve refractory materials. Special considerations are applicable to the analysis of the front half digestate due to complications related to the hydrofluoric acid.
The back half of the probe is a series of acidic aqueous solutions and potassium permanganate solutions. The back half probe is digested by the routine acid digestion techniques specified for EPA Method 7470A and EPA Method 200.8. The back half probe results in one analytical fraction for the ICP-MS metals and four analytical fractions for mercury.
MACT Metals Instrumental Method Techniques
Method 29 allows the metals to be analyzed by CVAA, GFAA, ICP-AES, and ICP-MS techniques. Mercury must be analyzed by a CVAA technique, but the other metals may be run by three different analytical techniques.
We use the ICP-MS technique (EPA Method 200.8) for most of the analyses. EPA Method 200.8 was chosen because it is able to simultaneously determine multiple metals cost effectively, is more sensitive for the MACT emissions criteria than ICP-AES techniques, and is a documented reliable method. The exception is the front half probe for the determination of arsenic and selenium. We can obtain lower reporting limits for arsenic and selenium when GFAA techniques are used on the front half of the probe. A comparison of the reporting limits for arsenic and selenium are shown in Table 6.
Calculation of MACT Metals Reporting Limits
Reporting limits in total micrograms are calculated based upon the final volume of the wet digestion procedure. On the front half of the sampling probe the final volume is determined by the wet digestion procedure. On the back half of the sampling probe, a final volume of 500 mL is used for the reporting limit calculations. Data used for the derivation of reporting limits shown in Table 4 can be found in Table 7.
The equivalent final volume is the final volume after digestion and any dilutions prior to the instrumental analysis. The reporting limit is calculated as shown below:
Reporting Limit in μg = (Equivalent Final Volume mL)(Instrument RL ng/mL)/1000
Hydrochloric Acid Sample Preparation Technique
In EPA Method 26, the hydrochloric acid emissions are captured in a 0.1 N sulfuric acid impinger solution. This strongly acidic solution is neutralized with a sodium hydroxide solution resulting in a two-fold dilution of the initial sample volume. The sample is then analyzed for chloride.
Chloride Instrumental Method Technique
All samples are analyzed in duplicate by ion chromatography techniques. The instrumental reporting limit is 0.1 μg/mL, which translates to 0.2 μg/mL in the original impinger solution. The chromatography can be challenging near the reporting limit because it is difficult to resolve low level chloride from background.
Calculation of Chloride Reporting Limits
The sample volume is diluted by a factor of 2 during the sample preparation techniques. With an instrumental reporting limit of 0.1 μg/mL, the resultant sample reporting limit is 0.2 μg/mL. Calculation of Total ug of chloride is derived by multiplying the impinger volume (mL) times the original impinger reporting limit (0.2 μg/mL).
Christian, Jeff. Columbia Analytical Services, Inc. Primary technical resource for analytical formulations andColumbia Analytical/Kelso SOPs.
Ferguson, Bruce 2005. Stack Testing for Boiler MACT, A Resource Guide. Prepared for the National Council for Air and Stream Improvement, Inc. (NCASI).
NCASI 2005. Boiler MACT Implementation, A Compilation of Questions and Answers, Version 2.0
National Emission Standards for Hazardous Air Pollutants for Industrial, Commercial and Institutional Boilers and Process Heaters, Subpart DDDDD of 40 CFR Part 63, §63.7480 through §63.7575. 2004 Federal Register
Proprietary Standard Operating Procedures of Columbia Analytical Services, Inc. 2003-2005.
United States. USEPA. Method 29 – Determination of Metals Emissions From Stationary Sources (Promulgated)
Tags: Chloride, EPA Method 1631, EPA Method 200.8, EPA Method 26, EPA Method 29, EPA Method 3050B, EPA Method 5050, EPA Method 6010, EPA Method 6020, EPA Method 7471, EPA Method 9056, MACT, Maximum Achievable Control Technology, Mercury, Reporting Limits, solid fuel, TSM