Accurate particle size distribution (PSD) data can provide a significant amount of information about the nature of a particulate emissions source. One means of collecting accurate particle size distribution data uses an in-situ cascade impactor. Figure 1 shows the actual samples from a well-performed in-situ cascade impactor test. Each photo shows particulate matter (the tiny dark spots) deposited on ultra-pure quartz substrates that occupy each of the stages of a cascade impactor. The particle size of the particulate collected in each stage gets successively smaller moving from the left to the right in the top photo (Photo 1) and continuing from left to right in the bottom left photo (Photo 2). The lower right photo (Photo 3) is the final backup filter and contains the smallest particles. Figure 1: A University of Washington Mark V Cascade Impactor Collected Sample Cumulative Mass Percentage PSD Data Plot The initial information obtained from analyzing the samples in these photos is a gravimetrically determined set of particle size cuts (D50) called a particle size distribution. When sampled from an industrial process, the distribution can be used to determine how much particulate matter of each particle size cut that a new piece of air pollution control (APC) equipment would need to remove from the gas stream to meet a target emissions value. For a controlled process, the distribution can tell us how efficiently an existing APC device is operating. Figure 2 shows a cumulative mass plot of a spline-fitted PSD data set. Note the X-axis title of “Aerodynamic Diameter.” Particle size distribution data are presented in terms of either the aerodynamic or the Stokes diameter of the particles. The aerodynamic diameter is a mathematically derived parameter that allows the irregularly shaped particles to be represented as idealized spheres of unit density. The Stokes diameter is a similar construct that represents the particles as spheres having the same density and settling velocity as the actual particles. It is important to know which diameter basis is being represented when evaluating PSD data. The “mass mean diameter (MMD)” is another parameter typically used to evaluate PSD data. The MMD is the particle diameter at which half of the sample mass is comprised of particles with smaller diameters. In the case represented in Figure 2, the MMD is approximately 2.5 microns. The smaller the MMD, the more difficult it is for an APC system to remove the particles from the gas stream. This trend tends to go up exponentially as particle size decreases. Once the MMD is below about 5 µm, a significant investment of money and energy will be required to remove a substantial fraction of the entrained particulate matter. Figure 2: Cumulative Mass Percentage PSD Data Plot dM/dLogD PSD Data Plot Another common way to represent PSD data is the use of a dM/dLogD graph. The dM/dLogD plot presents the mass concentration on a linear Y-axis versus particle size on a logarithmic X-axis. By using a linear scale for the Y-axis, the area under the curve between any two particle sizes equals the total concentration of particulate existing between the diameters. This type of plot is particularly useful because it provides insight into the type and prominence of formation mechanisms of the captured particulate matter. Figure 3 is a dM/dLogD plot that shows a trimodal distribution, with the three distinct peaks reflecting the three modes of particle formation from a combustion process. Figure 3: dM/dLogD PSD Data Plot The peak on the far-right side centered at 20 µm is typical of “mechanically” generated particles. These are usually caused by processes such as grinding or crushing. In the combustion case, they are the larger pieces of material that have not fully combusted and still retain a significant portion of combustible content. The peak can also represent agglomerations of finer particles that act aerodynamically as larger particles. The large center peak of approximately 1 µm that dominates the particle size distribution is that of the inert products of combustion. This mass of material is approximately 3½ times as prevalent in the gas stream as that represented by the other two peaks. A substantial portion of this material would need to be removed to effectively “control” particulate emissions from this source. The third, leftmost, peak centered at 0.1 µm is generated by particles that have been formed by vapor phase condensation. Gas phase compounds have cooled enough to form solid material in the gas stream. It is typical for these to form in the sub-micron size range. These particles, if present in a significant quantity, present a significant APC challenge. Comparison of Process Time Related PSD Data Sets Multiple dM/dLogD plots taken at different times can present a picture of the particle size evolution during a process. Figure 4 shows the graph of two data sets from the same process. Run 1 was collected during the initial few minutes of combustion while Run 2 was collected after the combustion had fully developed. The shifts in the sizes and relative magnitudes of the peaks discussed above are markedly apparent and are indicative of changes in the particle size generation mechanisms as the combustion advances. Figure 4: Comparison of Process Time Related PSD Data Sets SEM Analysis In addition to regular gravimetric analysis of cascade impactor samples, another analytical tool that can be employed is scanning electron microscopy (SEM). SEM analyses can provide information about the specific morphology and actual size of the collected particulate matter. This tool provides a visual means to distinguish between individual particles and agglomerations of smaller particles acting like a larger particle in the gas stream. Figure 5 shows an SEM photograph of a collected sample of coal fly ash. Figure 5: SEM Photo of Coal Fly Ash Cenospheres Size Specific Chemical Speciation using ICP-MS In addition to SEM analysis, a variety of other analytical protocols can be applied to the individual recovered substrates to get size-specific chemical speciation. This can be used to correlate one or more specific compounds in the gas stream to each particle size cut-point. This can then be related back to the particle generation mechanism for each cut-point. Figure 6 is a dM/dLogD plot of a metal coating process. After gravimetric determinations were made, each substrate was analyzed for a selection of target metals by ion coupled plasma - mass spectroscopy (ICP-MS). The result is a separate dM/dLogD plot for each target metal. As expected, the metals in the exhaust gas stream undergo vapor phase condensation to form submicron particles. These ultrafine particles present both a potential health risk as well as a significant challenge for emissions control. Figure 6: dM/dLogD plot Showing Size Specific Chemical Speciation using ICP-MS XRD Analysis of a Single Impactor Substrate Another tool is the use of X-ray diffraction (XRD) analysis of the collected sample on each substrate. XRD provides information about the crystallographic structure and chemical composition of the sample. Figure 7 is an XRD analysis of a substrate that was part of an impactor sample collected downstream of a venturi scrubber demister. This information was used to determine how much gypsum from the scrubber liquor was being lost because of subpar demister performance. Figure 7: XRD Analysis of a Single Impactor Substrate The examples presented herein demonstrate some of the information that can be generated from in-situ cascade impactor sampling. This information can help end-users assess the efficacy of their APC equipment or specify new equipment in a more informed and capital sensitive way. The author of this article was involved with the development of the Pilat, University of Washington impactor as well as other PSD tools. For further information on determining particle size distribution in any type of gas stream, contact email@example.com
I recently had a coworker ask about a conversation he’d had with a potential client. The client wanted to test for VOC (Volatile Organic Compounds). [Quick aside: I apologize for all the acronyms, but you cannot work in the world of stack testing without running into them, running over them, and occasionally being completely befuddled by one you’ve never seen.] He wanted to know what was “standard practice” for that kind of testing. I chuckled to myself; he was entering a world that was less decision tree and more decision bush, or decision labyrinth (with the local agency playing the role of minotaur). In order to properly test for VOCs, (at least) four questions need to be answered before determining what would be “standard practice.” Why are you looking for VOCs? What compounds are you looking for? What levels are you expecting to see? What levels do you need to see? What type of process are you going to test? Let’s look at each of these in more detail. 1. Why Are You Looking for VOCs? The primary reason for VOC testing is that EPA or another agency requires it through a permit. Permits often specify what methodology to use, what compounds to look for, and what the required detection limits are. Depending on the permit, a facility may be asked to provide emissions rates, emission factors, destruction efficiency, control efficiency, or something else entirely. A permit will often cite methodology, but there are often questions that remain. EPA Method 25A, a method geared towards the compliance world, may require a response factor and Method 18 is a decision tree in and of itself. A permit may define the methodology, but applying that method correctly is key. Beyond the regulatory drivers, VOC testing may be done to improve process efficiency, reduce fuel costs of an RTO (Regenerative Thermal Oxidizer), or to grab a baseline prior to installing a pollution control device. When testing is for non-regulatory reasons, a client may need real-time measurements, lower detection limits, or a more flexible testing approach. 2. What Compound are You Looking for? Surprisingly, clients rarely know the answer to this question. To properly select a VOC testing method, I need to know what I’m looking for and what type of gas stream it is in. Is it soluble? Is it sticky? Is it unstable? Is it deadly? Can it be collected in a summa canister, a Tedlar bag, or a sample train? Can I see it with GC (Gas Chromatography) or is the FTIR (Fourier-Transform Infrared Spectroscopy) or MS (Mass Spectroscopy) viable options? Different compounds can be detected using a variety of methods, knowing what is in a gas stream will improve the test program by helping to select the proper methodology for finding the compounds of interest. Alternatively, if this question remains unanswered, a pre-screening or evaluation of process input may provide valuable insight. 3. What Levels are You Expecting to See? What Levels Do You Need to See? Those two questions may seem the same (and often are), but there are good reasons to ask both questions. It is often acceptable just to know a compound is below a set limit without quantifying down to the ppt (parts per trillion) level when a permit is involved. However, destruction efficiency is hard to calculate if there are less than signs (non-detects) in both halves of the equation, in which case lower detection limits may be required. It’s worth noting that emission levels can change based on the process as well. Is there makeup air? Is it variable? Is the system steady-state? Is it continually venting? A changing gas flow rate can make quantifying certain compounds difficult. Knowing what detection limits are required to meet the permit requirements, design specifications, or engineering goals will help determine what VOC testing approach is appropriate. Agencies are asking for lower and lower emission measurement levels which means that traditional approaches may require modifications or fine tuning and opens the door for advanced sampling systems. 4. What Is the Process Being Tested? Will it be hot, pressurized, or ambient? Is there high moisture? What else is in the gas stream outside the compound(s) of interest? Will it interfere with the targeted compound? Is it deadly for the stack tester (this happens more than you’d think)? Is it variable with respect to time? Some methods provide a single data point for the duration of a run while others allow me to gather real-time data. Each of these answers will help to narrow down the VOC testing options and keep the test crew safe. If you’re confused by this, know that you are not alone. I know my co-worker was a little baffled, and probably still is. There are a lot of variables to sift through and questions to ask. There is not a one-size-fits-all solution to VOC testing. It is important that people performing these tests take the time to determine the proper sampling and analytical methodology to get appropriate detection limits, perform the test safely, and provide quality data that meets your needs. The author of this blog post has been in the stack testing field for over 15 years and the Team Leader for CleanAir’s Emission Measurement Group. For more information regarding VOC testing or other stack testing, contact Scott Lehmann at firstname.lastname@example.org.
Overview Determining the particle size distribution (PSD) of the filterable particulate matter (FPM) contained within a gas stream is often a necessary and valuable sampling tool. A PSD can be used diagnostically to evaluate the performance of an air pollution control device or to gather information to design a new one. It can be used to assess the nature of a product containing process gas stream with an eye towards recovery and re-use. In some instances, it can be used for compliance purposes by application of California Air Resources Board (CARB) Method 501. The most important decision to make is whether the PSD will be obtained using an in-situ device, which sizes the FPM as it is withdrawn from the gas stream at as-found conditions, or with a secondary method that evaluates the PSD of an extracted mass sample, usually in a laboratory. There are two primary reasons for doing extractive PSD work. The first, and most prevalent reason, is that the expertise to perform in-situ sampling is of limited extent in the industry. In-situ PSD sampling is technically challenging and is performed much less frequently than standard particulate matter source sampling methods. The second reason to use extractive sampling is that the gas stream cannot safely or reasonably be accessed with an in-situ device. A long-held particulate sampling tenet is, “Measure it while it’s moving.” The morphology and density of the particle have a strong effect on its behavior in the gas stream in addition to its actual physical size. The gas stream factors of temperature, pressure, velocity and moisture also come into play. With an extractive method, these factors are unavailable during the measurement process and can only be predicted to varying degrees of certainty. The two primary in-situ PSD devices are cyclones and cascade impactors. Cyclones The five-stage cyclone sampler developed by Southern Research Institute for the U.S. E.P.A is shown in Figure 1. Figure 1: EPA/SoRI Five-Stage Cyclone Sampler Each cyclone provides an individual particle size cut commonly referred to as a D50. The sizing range of the device shown is 10 to 1 micron. The sample is withdrawn isokinetically so the size distribution collected is representative of the PSD in the gas stream. Impactors Cascade impactors sample isokinetically and separate the sampled aerosol particles into size increments by inertial impaction of the particles on to a collection surface, which can be a lightweight greased steel foil, ultrapure quartz microfibre filter paper, Teflon™, or some other type of appropriate substrate material. This occurs at successive stages through the impactor, hence the name “cascade.” The Pilat, University of Washington, Mark V cascade impactor is shown in Figure 2. Figure 2: Pilat, University of Washington, Mark V Cascade Impactor Within this compact stainless-steel casing, 11 different D50’s ranging from ~20 to 0.1 microns are available. By changing stage configuration and adjusting sample rate the size distribution can be customized to a specific need. Cascade impactors are almost always used with a precutter head attached that enables sampling to be conducted in a standard, perpendicular to the gas stream orientation as shown in Figure 3. The precutter also can remove heavier loading of large sized particles that could overload the impactor. A special application for wet droplet PSD sampling is also possible with the precutter head. This application allows PSD sampling and analysis of saturated gas stream environments with droplets present such as downstream of a wet scrubber. Both droplet size and particulate size determinations are possible. The Pilat, University of Washington impactors have been considered the standard for in-situ PSD measurements and were the basis of the CARB 501 Method. Figure 3: Cascade impactor with precutter head The author of this blog post was involved with the development of the Pilat, University of Washington impactor as well as other PSD tools. For further information on determining particle size distribution in wet or dry gas streams, contact email@example.com.
I recently wrote an article on the long and painful history of condensable particulate measurement techniques for the Summer 2019 edition of the WPCA News, a publication of the Worldwide Pollution Control Association. The focus was on the “dilution method” currently referred to as OTM-37. I’ll hit a few of the highlights here but I recommend downloading the newsletter for the complete scoop. Condensable Particulate Matter (CPM) has had as long a regulatory history as its more density-endowed sibling Filterable Particulate Matter (FPM). In our recent podcast interview with Walt Smith, he reminisced about the good old days when EPA first proposed methods to measure particulate from a stack (Method 5) in 1977. In the proposed Method 5 rule, both FPM and CPM were included. In the final rule, CPM measurement was removed due to the many uncertainties involved with collecting the samples and interpreting the results. Have no fear though because CPM would have its day 15 years later with the promulgation of Method 202. EPA’s specific definition of CPM is… “Material that is vapor phase at stack conditions, but condenses and/or reacts upon cooling and dilution in the ambient air to form solid or liquid PM immediately after discharge from the stack. Note that all condensable PM is assumed to be in the PM2.5size fraction.”40 CFR 51 So gaseous compounds in the stack that turn into solid particles when they cool in the ambient air. Keep that in mind… ambient air. Method 202 was problematic from the get-go. Hot, wet stack gas was bubbled through big ice- cold test tubes filled with water… not exactly a good surrogate for ambient air. The Method was plagued with too many options and, most importantly, suffered from extremely high bias issues from false particulate created during the sampling process. Apparently, if you bubble reactive gases in a test tube… you get reactions. Who knew? To solve this problem, EPA drastically modified Method 202 to minimize contact between the water and the stack gas. In 2011 the “new” Method 202 was born. But the baby was still colicky. While the method modifications reduced the bias, they did not eliminate it – particularly where both SO2 and ammonia were present in the gas stream. Say… I don’t know… maybe downstream of an SCR. So remember the definition of CPM? Condenses or reacts in ambient air? Wouldn’t it be great if there were a way to simulate what actually happens to stack gas as it exits the stack and mixes and cools in the ambient air? Well friends, it turns out EPA has been working for years on just such a method. That’s OTM-37. It introduces cool, clean dilution air to the stack gas and provides mixing just like nature intended. Since this method mimics what actually happens to a gas plume in the ambient air, EPA has called this method the “gold standard” of particulate measurement. Everything seemed to be chugging along for Pinocchio to turn into a real boy… I mean for OTM-37 to be a real test method. Until one fateful day… Here’s where you download the WPCA News. Enjoy!
Walt Smith can rightly be called "The Father of Stack Testing." He's been involved in environmental monitoring since before there was an EPA. Walt was one of the original EPA employees and is responsible for authoring many of the fundamental stack test methods still in use today, including Methods 1 - 8. Recently, I had the pleasure of sitting down with Walt Smith for a chat about his remarkable 50+ years in the environmental monitoring field. We talked about his early days at EPA, how some of the stack methods testers use every day came into being, and his thoughts on the current state of emissions measurement. Currently, Walt offers training and consulting on a variety of emission measurement issues including QSTI training. You can contact Walt at waltersmith.com. Several of Walt's current presentations are available below. Also, his paper describing the nomograph he developed that he talks about in the interview is also linked below. How to Listen Part 1 of this interview is available on iTunes, Spotify, Stitcher, or wherever you get your podcasts. Part 2 will be released shortly. See the links below. Search for "Hot Air" or use one of the links below. Listen on the web: http://cleanair.libsyn.com Listen on iTunes: https://itunes.apple.com/us/podcast/hot-air-by-clean-air/id1413894085?mt=2 Listen on Spotify: https://open.spotify.com/show/24gUGGOvpygt3ayjAJtdbU?si=sbKPuLaVQFayMGPG50sEEQ Listen on Google Play: https://play.google.com/music/listen#/ps/Ivy2hniljmxncjhtj72jblrgcua Listen on Stitcher: https://www.stitcher.com/podcast/scott-evans/hot-air-by-clean-air?refid=stpr Recent Presentations from Walt Smith Click each link to view presentation. Pollutant Stratification Chain of Custody How to Get Good Data Stack Gas Simplified (Nomograph Paper)
In these inaugural episodes of CleanAir's podcast series Hot Air, I sit down with Jim Guenthoer to talk about the elements necessary for a successful stack test. Jim brings his 40+ years of stack testing experience to the table and provides lots of practical advice and anecdotes for both those who use stack testing services and those who provide them. Part 1 covers pre-test activities, Part 2 deals with the testing itself, and Part 3 addresses post-test activities. Our discussion is available as both an audio podcast and a video presentation. During the audio podcast, there are occasional references to the slides but not to the extent that it makes it difficult to follow without the slides. A link to download a copy of Jim's slides is provided below. The audio podcast can be heard on iTunes, Spotify, Stitcher, or Google Play or is available via a web link. All links are provided below. Subscribe to the podcast to receive new episodes of interest to anyone involved with air quality issues. Enjoy! The Successful Stack Test: Part 1: Pre-Test Why We Test Selecting the Test Contractor The Test Plan Choosing a Test Method Safety Site Access Site Support The Successful Stack Test: Part 2: The Test Equipment Mobilization and Setup Sampling and Data Collection Sample Recovery Wrapping Up Activities The Successful Stack Test: Part 3: Post-Test Sample Analysis and Laboratories Test Report Considerations Concluding Thoughts You can listen and subscribe to Hot Air on iTunes, Google Play, Spotify, or anywhere you get your podcasts. The Successful Stack Test Listen on the web: http://cleanair.libsyn.com Listen on iTunes: https://itunes.apple.com/us/podcast/hot-air-by-clean-air/id1413894085?mt=2 Listen on Spotify: https://open.spotify.com/show/24gUGGOvpygt3ayjAJtdbU?si=sbKPuLaVQFayMGPG50sEEQ Listen on Google Play: https://play.google.com/music/listen#/ps/Ivy2hniljmxncjhtj72jblrgcua Listen on Stitcher: https://www.stitcher.com/podcast/scott-evans/hot-air-by-clean-air?refid=stpr Watch the Video Slide Presentation: https://vimeo.com/album/5444815 Link to a copy of Jim's slides: https://cdn2.hubspot.net/hubfs/3331037/Guenthoer%20P6.pdf
1. Hydrogen Chloride (HCl) is an acid gas classified as a Hazardous Air Pollutant (HAP), and is identified as harmful to the environment and human health. HCl is highly corrosive and can damage metal structures over time. It is also highly water soluble (known as Hydrochloric Acid in solution) and will affect the chemistry and ecology of bodies of water or certain types of soils. 2.Hydrogen Chloride regulation. The U.S. EPA regulates HCl across several industries, including fossil-fueled power plants, refineries, cement kilns, pharmaceutical manufacturers, and steel manufacturers. Within the U.S., HCl emissions must be measured, estimated, or calculated, and periodically reported to regulatory agencies. If an affected source measures HCl directly, then the source owner must use a specified sampling methodology dictated by the regulation affecting the source. 3. US EPA Method 26. Originally promulgated in 1991, this test method extracts a sample from the stack or duct at a constant rate and passes the sample through a filter to remove particulate. After the filter, reagent-filled impingers capture chloride ions as the sample is bubbled through the sampling train. Since the method does not measure HCl directly, but rather the captured chloride ions, other chlorine compounds have the potential to generate a positive bias, if present. The chloride ion concentration is measured via ion chromatography. 4. US EPA Method 26A. Originally promulgated in 1994, this method employs the same underlying principles as Method 26 while conforming to most Method 5 specifications for sampling particulates. Specifically, Method 26A isokinetically extracts a sample from the stack or duct and passes it through a Method 5 filter to remove particulate. The impingers capture chloride ions just as in Method 26; however, to account for higher flow rates and higher sample moisture levels, the impingers hold larger volumes than in Method 26. The EPA created this method to address possible sampling biases in “wet” stacks, or stacks that are at or near the moisture saturation point and where entrained water droplets containing HCl may be present. 5. US EPA Method 320. Method 320 is a spectrographic method. Unlike the “wet” Methods 26 and 26A, this method measures the HCl in the gas phase without first absorbing it into a liquid reagent. A sample is extracted from the stack or duct at a constant rate and passed through an FTIR sample cell. The FTIR scans the volume of gas in the cell and measures HCl (as well as other compounds) in near real-time. American Society for Testing and Materials (ASTM) D6348 12e1 is an FTIR method like Method 320 that may also be used when FTIR is approved for use at a source. These methods directly measure HCl, not chloride ions and so are less susceptible to some biases. 6. Which method should be used for compliance? The Mercury and Air Toxics Standards (MATS) Rule lowered the HCl emissions limits for all existing power generating units, while giving sources different options for complying with the new limits. Certain sources may use sulfur dioxide (SO2) as a surrogate for HCl, certain low emitting sources (LEE) may be exempt from monitoring HCl, and other sources may choose to comply with a permanent HCl CEMS. For sources not qualifying for or opting not to use these compliance options, Methods 26, 26A, and 320 are the HCl measurement methods to be used for quarterly stack testing compliance demonstration. For sources with a wet stack, Method 26A must be used. For sources with a dry stack, Method 26 or Method 320 can also be used. Non-power generating units historically restricted methodology to Methods 26 and 26A for HCl compliance; however, FTIR has been approved for some sources in some states. 7. What about measurements that are not for compliance? Recently, source owners have installed dry sorbent injection (DSI) and other emission control devices to comply with lower HCl limits. During the performance guarantee for newly installed control devices, as well as subsequent tuning or diagnostic testing, source owners choose the measurement methodology that they believe best helps them meet their test program goals and quality objectives. There are no regulations that specify measurement methodology when the testing is not for compliance purposes. 8. Entrained water biases in Method 26. Method 26 uses small filters, small impingers, and extracts samples at a slow, constant rate. When sampling a wet stack, this method will not representatively sample entrained water droplets. Typically, the low flow rate and small sampling train makes it likely that smaller entrained water droplets and gas molecules will be overrepresented in the sampling train, while larger entrained water droplets will be underrepresented in the sampling train. This means that the HCl concentration will typically be biased low when Method 26 is used in the presence of entrained water droplets. 9. Chlorine biases. Chloride salts and elemental chlorine exist in the flue gas and cause molecular interactions to occur in the bulk flow. Conditions found in typical flue gases drive conversion of most chlorine content to gaseous HCl. However, any residual chlorine in the bulk flow could also convert to HCl in the sampling system if the sampling system is not carefully constructed and operated. This transport issue can be a problem with all of the methods referenced above. 10. Keep it hot, rinse a lot. Chloride ions deposit on the surfaces of unheated and chilled portions of the sampling train (the impingers), so the post-run collection and rinse methodology for wet methods becomes very important. Sample system temperature has a direct effect on the degree of chloride deposition. Sample system temperature must be kept as hot as possible during sampling at all points. Any cold spots, even small ones, can introduce a significant bias to the measurement. Also, laboratory studies show that a large volume of rinse water and thorough wetting of all surfaces of sampling equipment is necessary to fully remove all collected chloride. The important of thorough rinsing cannot be overstated – research shows that recovery and rinsing bias is tester-specific. Like the wet methods, Method 320 also requires transport of the sample from the source to the instrument. Adequate and consistent heating of the gas sample along the entire transport path is critical in achieving good performance with this method. M26 and M26A specify a temperature range of 248-273 °F, while M320 does not specify temperature. Temperatures must not exceed 400°F if Teflon components are used since Teflon is unstable above this temperature. 11. What are some alternatives? There have been mixed results with in-stack measurements using FTIR as well as tunable diode laser (TDL) technology. These approaches eliminate sample transport biases entirely, but are more amenable to long-term measurements and CEMS applications rather than short-term compliance tests. More recently, EPA has considered the use of sorbent traps (OTM-40) for HCl as Alternate Test Method 129 (ALT-129), which can eliminate some of the transport concerns. Overall, HCl is a difficult compound to quantify, not because of a lack of analytical techniques that can measure HCl, but because of the complications of sample transport. CleanAir conducted a study, for the Electric Power Research Institute (EPRI) comparing wet methods to FTIR for halide measurements. Also, in the special case of a supersaturated stack, CleanAir performed diagnostic testing to help troubleshoot HCl issues with an isokinetic FTIR. If you have any questions concerning HCl measurement, we can help you with research, data, and testing.
The U.S. EPA recently “upgraded” Other Test Method 40 (OTM-40) to an approved alternative method (ALT-129) for coal-fired Electric Generating Units (EGUs) subject to 40 CFR Part 63 Subpart UUUUU, otherwise known as the Mercury and Air Toxics Standards or MATS. The change from OTM to ALT means the method can now be used by coal-fired EGUs without pre-approval . The method is the same, just the broad approval is new. The approval does come with some restrictions, however. The method may only be used without pre-approval by coal-fired EGUs and only those with low moisture flue gas at temperatures above 212 °F. The flue gas can contain no entrained droplets. Under these conditions, field testing shows OTM-40 performs similarly to Method 26A (M26A, one of the traditional HCl Reference Methods) but could potentially be biased high. A simpler sample train setup, lack of harmful chemical reagents, and quicker sample recovery may contribute to lower testing costs for affected facilities compared to Method 26 or 26A. OTM-40 uses HCl sorbent traps and the same equipment and procedures as Method 30B (M30B, mercury sorbent trap method). The sample gas flows through paired sorbent tubes which capture HCl as chloride. Sample recovery and analysis typically consist of a water extraction followed by ion chromatography. Like M30B, method performance is evaluated for each test using NIST-traceable spiking in one of the traps. This paired-sample design helps validate test results and assess method precision which is difficult to accomplish with the traditional test methods. While M26 uses a set of glass impingers filled with reagent to collect HCl outside of the stack, OTM-40 traps capture the HCl sample in-situ. This difference eliminates potential low bias resulting from insufficient heating of the sampling system, since the sample is collected in the traps before the sample gas is transported through the rest of the sampling equipment. However, this approach is susceptible to the same potential high bias as M26 from the presence of elemental chlorine (Cl2) or other chlorinated compounds in the flue gas. Unlike M26 or M26A, the trap method could also have a high bias if there are any chloride compounds contained in the particulate matter collected on the glass wool plug at the front of the trap. The trap method requires that the glass wool be included in the sample analyses, whereas the particulate filters in the conventional methods are not required to be analyzed. As mentioned above, OTM-40 is currently only approved for coal-fired EGUs with dry stacks. Stacks with wet scrubbers still require isokinetic sampling and nozzles for representative sampling of possible entrained water droplets. OTM-40 does not sample in this fashion. Also, the method has only been validated on coal-fired utility stacks; therefore, OTM-40 has not been approved for other sources although it may still be used with pre-approval. Click here for the approval letter, and here for the full method description.
Particulate is defined by the method used to measure it. Unlike chemically distinct emissions like carbon monoxide or nitrogen oxide, “particulate” can be composed of many different distinct compounds. Some of these compounds may be volatile and may be “particulate” at lower temperatures but gaseous at higher temperatures. US EPA Method 5. This test method isokinetically extracts a sample from the stack or duct and passes that sample through a filter. The sample probe and filter are heated to 248 °F. Any particulate captured on the filter is (logically) referred to as “filterable particulate” or FPM; that is, any material not gaseous at that temperature is captured. There are variants of Method 5 (for example, 5B and 5F) that use other temperatures for sample collection. US EPA Method 202. The gaseous material sampled with Method 5 passes through the filter and a temperature-controlled condenser followed by two, initially dry, impingers – essentially big test tubes. As the gas is cooled, water and other compounds (both organic and inorganic) condense out. Any material collected in these impingers (minus the water) is referred to as “condensable particulate” or CPM. The logic used for this method is that any condensed material would also condense and form particulate when the gases exit the stack and are cooled in the ambient air. More on this assumption in a moment... Method 202 Bias Issues. Unfortunately, there are often other compounds present in the sample gas that may react when placed together in a large, wet impinger. Two compounds commonly found in some gas streams are ammonia (NH3) and sulfur dioxide (SO2). Left on their own in the stack, these two gases will not normally combine even upon exiting the stack. However, when mixed together in an impinger with even a small amount of water, the two gases combine to form ammonium salts – a particulate. The formation of ammonium salts is an artifact of Method 202 sampling and does not occur with gas exiting the stack. Therefore, Method 202, can be biased high for gases streams containing these two compounds -- the method creates “false particulate.” A Synergistic Effect. NH3is highly water soluble. During a test, as NH3is dissolved in the impinger water, the pH of the water is raised. The higher pH enhances the collection of SO2. -- in effect, a little caustic scrubber is created. The higher the pH, the more efficiently SO2is scrubbed. The more NH3and SO2are captured, the higher the Method 202 bias. Some may say that when Method 202 was modified a few years back, it fixed the bias problem. Unfortunately, there is ample evidence that while the bias issues are better than they were with the old method, they are still with us. The addition of the “dry” impingers and CPM filter in the new method helped but did not completely solve the bias problem. What to Do? For facilities where highly biased CPM results are an issue, here is an approach that can work. For Method 202 analysis, results are split into organic and inorganic fractions. It is the inorganic fraction that typically is biased. It is where the ammonium salts are found. So the solution? Use only the organic fraction. But wait, I hear you say. What about any other inorganic condensable compounds that may be legitimately considered particulates? We can’t just throw away all the inorganics, can we? Controlled Condensation. In any gas stream that contains SO2, particularly those with catalysts to control NOx, there will be some SO3/H2SO4– sulfuric acid. At temperatures well above ambient, sulfuric acid will form a mist which is legitimately considered a particulate. At Method 5 collection temperatures (248 °F) Much of this mist will be captured on the filter, but some of it will make it through the filter and condense in the impingers. For many gas streams containing SO2, SO3will be the only significant contributor to CPM. This SO3must be accounted for since it is a legit particulate. The answer is to do a test that is specific to sulfuric acid mist… controlled condensation (CCM). Why Controlled Condensation? The CCM method is designed to condense SO3/H2SO4while minimizing water condensation. Method 202 condenses both water and acid gas simultaneously leading to the chemical interactions described above. While CCM does not completely eliminate the potential for the formation of false particulate, it performs much better than M202 in this regard. An unbiased CPM measurement. So a less biased measurement of CPM can be obtained by using the results of the CCM test in place of the Method 202 inorganic fraction. The total CPM would then be: CPM = Method 202 organic fraction + CCM results Both tests should be conducted either simultaneously or as close as possible in time. In order to use this approach, regulatory authority permission must typically be granted. CleanAir has a white paper available on this issue with more detailed information, data, and a case study. If you are interested in this approach to unbiased CPM measurement, we can support you with research, data, and testing. Let us know if we can help.
Performance Specification 11 (PS-11) establishes the initial installation and performance procedures required for evaluating the acceptability of a particulate matter (PM) continuous emission monitoring system (CEMS). Multiple industries utilize PM CEMS and are influenced by applicable EPA regulations; such as the Portland Cement MACT, Industrial Boiler MACT, and Utility MACT. Evaluation of the performance of a PM CEMS over an extended period of time, or to identify specific calibration techniques and procedures to assess their performance is covered under Procedure 2 of Appendix F—Quality Assurance Requirements for Particulate Matter Continuous Emission Monitoring Systems Used at Stationary Sources. In all there are hundreds of pages (and a few spreadsheets too!) that must be reviewed to become familiar with PS-11 and PM CEMS requirements. Newly installed PM CEMs operations will bring about a multitude of difficult questions. Correlation curves and coefficients, RRA, RCA, ACA, SVA, precision, bias, stratification, regression analysis, what does it all mean? Do they apply to my current test program? How do I operate to perform the tests at three different PM concentration or loading levels? Should I perform EPA Method 5, 5B or MATS 5? Do I need to worry about condensable particulate matter (CPM)? How does PS-11 influence my future PM Relative Accuracy Test Audits (RATAS)? What is the method detection limit (MDL) of the stack test? How do I know good reference method (RM) test data from bad? Can the tester perform overlapping (staggered) test runs to obtain data points more quickly? Should I conduct paired, otherwise known as collocated, reference method test runs? CleanAir has done the hard work for you and summarized the important information you need to know in CleanAir “Layperson’s Guide to PS-11”.
CleanAir is now accredited to ISO 17025, the international standard for competence of testing laboratories. This is in addition to two other accreditations we hold: ASTM D7036 (Standard Practice for Competence of Air Emission Testing Bodies) and the NELAC FSMO standard for Field Activities. In order to obtain these accreditations, CleanAir underwent rigorous third party quality audits both in our offices and in the field. These credentials provide assurance to our clients that CleanAir’s quality system and practices meet stringent US and international quality standards. Scope of Accreditation For more detailed information and for copies of our accreditation certificates, please see our Accreditation page.
If your facility is an EGU that complies with Mercury and Air Toxics Standards (MATS), you are probably aware of some new electronic reporting dates in the coming year. On April 6, 2017, a final rule on MATS electronic reporting was published in the Federal Register (82 FR 16736) that revised Reporting Requirements (40 CFR 63.10031). Under the final rule, performance evaluation tests (RATA, RCA, RRA, etc.), PM CEMS quarterly reports (or the approved alternatives), and compliance reports required by 40 CFR Subpart UUUUU must be reported as a portable document format (PDF) using the Emissions Collection and Monitoring Plan System (ECMPS) Client Tool. This reporting requirement is in place until July 1, 2018, at which time these reports must be uploaded to the Central Data Exchange (CDX) via the Compliance and Emissions Data Reporting Interface (CEDRI). The Electronic Reporting Tool (ERT) is a software program that is used to put your data into a proper format to be uploaded in CEDRI. Essentially, your regulatory requirements pertaining to MATS will not change this year; however, your regulatory reporting requirements will change. These new requirements are part of EPA’s Cross-Media Electronic Reporting Rule (CROMERR), which spans across all EPA regulatory programs and has two primary functions: to provide a streamlined, uniform electronic approval process; and to aggregate data in a central database that is available to the public. If you have not encountered the ERT, CEDRI, and/or CDX, it is possible that your testing contractor or consultants have dealt with the process and can guide you through this process. In fact, there is provision in the process where your testing contractor can do most of the data entry and submission preparation. If you have any questions, comments, or experiences you wish to share regarding any aspect of electronic reporting, feel free to leave a comment below, or reach me directly at firstname.lastname@example.org. Happy Holidays!
What do EPA's proposed changes to Method 202 mean for someone who is hiring a contractor to perform stack emission sampling? In late August of 2017, EPA proposed revisions to Method 202 for condensable particulate matter (CPM). EPA states that its goal with the proposed revisions is to improve consistency in the method and eliminate known biases. You can see their summary page here. These proposed changes will impact stationary source owners and operators that must use Method 202 to determine compliance with standards regulating particulate matter smaller than 2.5 microns (PM2.5), or any stationary source owner/operator that is required to measure CPM via this method for any reason. What changes is EPA proposing to make? There are numerous proposed changes to the method, but they typically fall into four categories. Eliminating Systemic Errors – These changes attempt to limit the biases that the method currently introduces to the measurement. Unfortunately for operators of these sources, many of the biases work against you. Revisions here include modifications to the sample train, suggested limits to run times, and increases to the requirements for calibration recordkeeping. Limiting Optional Equipment or Procedures – To increase industry-wide consistency of the measurement, EPA is limiting some of the choices that testers have historically been allowed to make. This includes changes to the rules regarding field train blanks and field reagent blanks, and clarifications on the applicability of the method. Quantifying and Reducing Blank Contamination – The EPA is proposing modifications to glassware cleaning procedures, reagent blanks, and blank correction calculations. These changes are intended to more appropriately quantify the systemic bias in the method. These changes are important to plants since any residue in the sampling equipment in excess of a specified limit is considered to be CPM from your process. Adding Flexibility – Counter to the above categories, EPA is proposing to allow some options that the testing industry has shown to be advantageous. Stack testers will be allowed more options for sample containers and weighing containers under these proposed changes. To see all the changes in detail, visit EPA’s website. What will change when you bring a contractor onsite to conduct a Method 202 test? From an operator’s perspective, not much will change. The method will remain mostly unchanged and this is all part of an improvement process that dates back for many years. Someone with a keen eye might notice a few things though including: A longer setup time. Vertical condensers in the sample train. Run times limited to 2 hours. Also, the report will change slightly. It will take someone with an even keener eye to notice: A report with more ancillary data regarding the train purge, balance calibrations, and other recordkeeping. Higher allowable blank reductions. Slightly lower CPM numbers due to elimination of biases. What is the timeline for comments? The comment period lasts for 60 days. We are already more than halfway through that; the comment period ends in late October. As for when the proposed rule changes will go into effect, that is unknown at this time. Will my testing contractor be ready for all this? It's hard to say. CleanAir's stack testing team, like many testing organizations, has already incorporated many of these changes into our procedures based on a 2016 guidance document issued by EPA. Hopefully, most of the changes can be implemented without incurring much in the way of additional costs.
A successful stack test begins long before your favorite stack testing company arrives at your facility. In fact, a successful stack test begins before you choose a stack testing company! When you are planning for your test, the first thing to do is define the objectives of your test. In other words, identify WHY you are doing this test. Working With Your Stack Tester A three-run compliance test is easy. But a multi-method diagnostic test may not be. Then, be sure your stack tester understands the objectives as well. The better they understand what you are trying to accomplish, the better prepared they can be to help you achieve your objective. Once you have communicated the project objectives to the stack testers, let them figure out HOW the work is going to get done – what methods and sampling strategies will provide you with the best data. They are the experts! Let them do that work for you. Once you’ve chosen your testing company and have awarded a contract, it’s time to start coordinating how this work is going to be done. Often, this requires a site visit by your stack tester. Other times, this can all be done remotely. It’s important to iron out exactly when and how the test is going to be executed. When is safety training offered at your facility? Who is responsible for providing power? Tasks such as safety training and hooking up power can take almost as long as the actual tests and not being ready can lead to costly delays. Preparing Your ERT Additionally, US EPA has recently expanded (and will continue to expand) requirements related to the Electronic Reporting Tool (ERT). The ERT is used by facilities to report the results of compliance emission tests to both state and federal regulatory agencies. Most stack testing companies will prepare your ERT submission file for you, but you may need to provide some additional information to them for the submission to go smoothly. It may sound cliché, but you’ve probably noticed that communication is key. Keeping an open line of communication with your stack tester when you are soliciting proposals, when you are planning for a test, when you are executing a test, and when you are reporting results can make your life much easier and result in a well-executed test. CleanAir has prepared a general checklist for conducting a successful stack test.