A CEMS is a comprehensive system of equipment used to continuously measure, analyze, record, and report gaseous and particulate matter emissions from stationary sources, such as power plants, manufacturing facilities, and industrial boilers. Its primary purpose is to ensure compliance with environmental regulations set by agencies like the EPA.

The system is comprised of several key parts:

• Sample Probe: Extracts a representative sample of the flue gas.
• Sample Transport/Conditioning System: Moves the sample to the analyzers and prepares it by removing moisture, particulates, or interferences (in extractive systems).
• Analyzers: The core instruments that measure the concentration of specific pollutants (e.g., SO₂, NOx, CO, O₂).
• Opacity Monitor (Transmissometer): Measures the particulate matter density by projecting a light beam across the stack.
• Data Acquisition and Handling System (DAHS): A computer-based system that collects, processes, stores, and formats the data from the analyzers into reports required by regulatory bodies.

CEMS are required primarily for regulatory compliance. They provide a continuous and verifiable record of emissions, which is used for:

1. Regulatory Reporting: Facilities must report their emissions to government agencies (like the EPA in the U.S.) to demonstrate they are operating within the limits set by their permits (e.g., under the Clean Air Act).
2. Process Control: While its main function is compliance, CEMS data can also be used by plant operators to optimize combustion processes, improve efficiency, and reduce pollution control costs. For example, high NOx or CO readings can indicate poor combustion.
3. Emissions Trading: In market-based programs (like the Acid Rain Program), CEMS data is used to quantify the "allowances" or "credits" a facility has used or can sell.
4. Public Transparency: The data provides accountability and assures the public that industrial facilities are being monitored for their environmental impact.

A CEMS is an integrated system, and its components work together. The main parts include:

• Sample Interface: The probe and filter assembly that extracts the gas sample from the stack or duct. It's the first point of contact with the flue gas.
• Sample Conditioning System: (For extractive systems) This part "cleans" the sample. It typically includes a sample line (often heated to prevent condensation), a gas cooler/chiller (to remove water vapor), moisture sensors, pumps, and additional filters.
• Calibration Gas System: A system to introduce known concentrations of "span" (high-level) and "zero" (clean air or nitrogen) gases to the analyzers to check their accuracy and perform adjustments.
• Gas Analyzers: A suite of instruments, each designed to measure a specific gas (e.t., an NDIR analyzer for CO, a Chemiluminescence analyzer for NOx). This also includes the diluent gas analyzer (O₂ or CO₂), which is necessary to correct emission concentrations to a standard level.
• Opacity Monitor: A separate instrument (transmissometer) that measures particulate matter by shining a light beam across the stack.
• Data Acquisition and Handling System (DAHS): The "brain" of the CEMS. It's a computer and software package that polls the analyzers for data, performs calculations, validates data, generates alarms, and creates the final emission reports in the specific format required by the regulatory agency.

This is a fundamental distinction in CEMS technology, based on *where* the gas measurement takes place:

• Extractive CEMS:
  • This is the most common type. It *extracts* a sample of the flue gas from the stack using a probe.
  • The sample is then transported via a heated sample line to a remote analysis shelter or cabinet, which houses the analyzers.
  • Before analysis, the sample *must* be conditioned (cooled to remove moisture, filtered to remove particulates).
  • Pros: Analyzers are in a clean, climate-controlled environment, making maintenance easier. Can measure multiple gases with one system.
  • Cons: Complex sample conditioning system is prone to failure (clogging, leaks, pump failure). Time lag between sample extraction and analysis.
• In-Situ CEMS:
  • This type measures the flue gas *inside* the stack. The analyzer is mounted directly on the stack or duct.
  • There is no sample extraction or conditioning system.
  • Pros: Real-time measurement with no time lag. No complex sample system to maintain. Measures the gas in its true state.
  • Cons: Analyzers are exposed to the harsh stack environment (high temperature, vibration, corrosive gases). Maintenance and calibration can be more difficult (requires accessing the stack).

This is a further breakdown of *extractive* CEMS, based on how they handle moisture in the sample gas:

1. Cold-Dry Extractive CEMS:
   • This is the traditional method.
   • The sample is extracted and then *cooled* using a thermoelectric chiller (or similar device) to a low, stable temperature (e.g., 4°C).
   • This condenses the water vapor, which is then removed by a peristaltic pump.
   • The resulting "cold, dry" sample gas is then sent to the analyzers.
   • Why do this? Most conventional analyzers are not designed to handle high moisture, and water can interfere with measurements.
   • Note: Since water is removed, the measured concentrations are higher than in the stack. The DAHS must use data from a moisture sensor (or a calculation) to correct the readings back to a "wet basis" for reporting.
2. Hot-Wet Extractive CEMS:
   • This method keeps the *entire* sample system hot—from the probe tip to the analyzer.
   • The probe, sample line, and even the analyzer cell are heated to a temperature *above* the dew point of the flue gas (e.g., 180°C).
   • This prevents *any* moisture from condensing. The gas is analyzed on a "hot, wet" basis, which is its true state in the stack.
   • Why do this? It avoids the complexity of the sample cooler. It's also necessary for measuring water-soluble gases (like ammonia or HCl) that would be "scrubbed out" and lost if the sample were condensed.
   • Note: The analyzers used must be specifically designed to operate at high temperatures.

The DAHS is the computer and software component of the CEMS. It is just as critical as the hardware, as it's responsible for turning raw analyzer signals into a compliant emissions report. Its key functions are:

• Data Collection: Continuously polls the analyzers (e.g., every 1-10 seconds) to get the latest concentration readings (ppm, %, etc.).
• Data Averaging: Calculates the 1-minute, 15-minute, and 1-hour averages required for compliance reporting.
• Calculations:
  • Performs mass emission rate calculations (e.g., lbs/hr) using flow meter data.
  • Corrects emissions to standard O₂ or CO₂ levels (e.g., "ppm corrected to 7% O₂").
  • Corrects "cold-dry" readings back to a "wet basis."
• Data Validation: Automatically checks the status of the CEMS. It flags or invalidates data that is collected during periods of calibration, maintenance, or analyzer malfunction. This ensures only valid data is used in compliance reports.
• Alarming: Generates alarms for operators if emissions exceed a set limit, if the CEMS fails a self-check (like a daily calibration), or if a component (like a pump) fails.
• Reporting: Automatically generates the detailed electronic reports (EDRs) and hardcopy summaries in the exact format required by the regulatory agency (e.g., EPA's Part 75 or Part 60).

• ppm (parts per million): This is a unit of concentration. It means "one part of a substance for every one million parts of the total mixture." For gases, it's a volume-to-volume ratio. For example, 100 ppm of CO means that in one million "parts" (e.g., liters) of flue gas, 100 of those "parts" are CO.
• Corrected to 7% O₂: This is a "normalized" concentration. Flue gas is mostly nitrogen and CO₂, with varying amounts of excess oxygen (O₂) left over from combustion.
  • If you add more excess air (called dilution air) to the process, you don't change the *total amount* of pollutant being emitted, but you *do* dilute it. The ppm reading will go down, making it seem like you are polluting less, which isn't true.
  • To prevent this, regulators require emissions to be reported at a standard O₂ level (e.g., 3%, 6%, or 7% O₂).
  • The DAHS uses a mathematical formula with the *measured* O₂ level and the *measured* pollutant level to calculate what the pollutant concentration *would be* if the O₂ level were exactly 7%. This allows for a fair, "apples-to-apples" comparison between different sources.

The O₂ analyzer is a critical "diluent" monitor. It typically doesn't measure a pollutant, but its measurement is essential for two reasons:

1. Emission Rate Normalization: As described in the previous question, the O₂ reading is required to normalize or "correct" the pollutant concentrations (SOx, NOx, CO) to a standard reference level. This is a regulatory requirement to ensure emission data is comparable.
2. Combustion Efficiency Monitoring: From an operational standpoint, the O₂ level is a key indicator of combustion efficiency.
   • Too high O₂: Indicates too much excess air is being fed to the boiler. This wastes fuel, as the extra air just gets heated and sent up the stack, carrying energy with it.
   • Too low O₂: Indicates incomplete combustion. This is inefficient (not all fuel is burned) and dangerous, as it leads to the formation of high levels of CO (a poison) and soot.
   Operators use the O₂ reading to trim the air/fuel mixture for optimal performance.

A "representative sample" is a sample of flue gas whose chemical composition and properties are as close as possible to the *average* composition of the entire gas stream in the stack at that moment.

It's critical because a measurement is useless if it's not representative. For example, measuring only at the stack wall might give a reading biased by air leaks or cooler temperatures.

To ensure a representative sample, several things are done:

• Probe Location: Regulations (like 40 CFR Part 60, Appendix A, Method 1) specify exactly *where* to locate the CEMS probe. It must be in a section of straight duct, far downstream from flow disturbances (like bends or dampers) and far upstream from the stack exit. This ensures the gases are well-mixed.
• Probe Type: A simple, single-hole probe is often sufficient for well-mixed gas streams. For large-diameter stacks, a "traversing" probe or a multi-hole (averaging) probe might be used to sample from several points across the duct's cross-section.
• Filtering: The probe must have a filter at its tip to prevent "stratification" of particulates, keeping ash and soot out of the sample line while allowing the gas to pass through.

In the United States, CEMS requirements are primarily driven by the Environmental Protection Agency (EPA) under the Clean Air Act. The key regulations include:

• 40 CFR Part 75: This is the "Acid Rain Program" regulation. It's one of the most stringent CEMS regulations and applies to coal-fired power plants. It mandates continuous monitoring of SO₂, NOx, CO₂, and stack gas flow rate to calculate mass emissions (tons/year) for the emissions trading program.
• 40 CFR Part 60: These are the "New Source Performance Standards" (NSPS). This regulation sets emission limits for *new* or *modified* industrial sources. Various subparts specify CEMS requirements for different industries (e.g., Subpart Da for large boilers, Subpart J for refineries). It often requires CEMS for SO₂, NOx, CO, and Opacity.
• 40 CFR Part 63: These are the "National Emission Standards for Hazardous Air Pollutants" (NESHAP), often called the MACT (Maximum Achievable Control Technology) standards. These rules target specific hazardous air pollutants (HAPs) and may require CEMS for pollutants like HCl, mercury, or total hydrocarbons.
• State and Local Permits: Individual facilities also have operating permits issued by their state or local air quality agency. These permits often incorporate federal rules and may add their own, more stringent CEMS monitoring requirements.

• What it is: SOx refers to sulfur oxides, primarily Sulfur Dioxide (SO₂). Sulfur Trioxide (SO₃) also exists but in much smaller quantities. SO₂ is a colorless gas with a sharp, pungent odor.
• Why it's monitored:
  1. Acid Rain: SO₂ is the primary cause of acid rain. It reacts with water and oxygen in the atmosphere to form sulfuric acid, which damages forests, lakes, and buildings.
  2. Human Health: SO₂ is a respiratory irritant. It can aggravate asthma and other respiratory conditions, leading to hospitalizations and premature death.
  3. Source: It's formed when fuel containing sulfur (like coal and oil) is burned. CEMS for SO₂ are almost universal on coal-fired power plants and oil refineries.

There are two primary methods used in modern CEMS:

1. Non-Dispersive Infrared (NDIR):
   • Principle: This is a very common method. It relies on the fact that SO₂ molecules absorb infrared (IR) light at a very specific wavelength.
   • How it works: An IR light source is shined through a sample cell containing the flue gas. On the other side, a detector measures how much of that *specific wavelength* of light was absorbed. The amount of light absorbed is directly proportional to the concentration of SO₂ in the cell.
   • Interference: A "correlation wheel" or optical filter is used to isolate the SO₂-absorbing wavelength and a "reference" wavelength, which helps correct for interferences from other gases (like CO₂) or dust on the cell windows.
2. Non-Dispersive Ultraviolet (NDUV) / UV Fluorescence:
   • Principle: This method uses ultraviolet (UV) light.
   • How it works (Absorption): Similar to NDIR, but it measures the absorption of UV light at a wavelength specific to SO₂. This is often used in in-situ CEMS.
   • How it works (Fluorescence): A UV lamp excites the SO₂ molecules in the sample. The excited molecules then "fluoresce," or release that energy as light at a *different*, specific wavelength. A detector measures this fluorescent light, which is proportional to the SO₂ concentration. This method is very sensitive and has few interferences.

• What it is: NOx (Nitrogen Oxides) is a blanket term for a family of gases, primarily Nitric Oxide (NO) and Nitrogen Dioxide (NO₂). In combustion, NOx is formed mostly as NO (around 90-95%), which then slowly oxidizes in the atmosphere to form NO₂.
• Why it's monitored:
  1. Ground-Level Ozone (Smog): NOx is a primary ingredient in the formation of ground-level ozone (smog). NOx reacts with volatile organic compounds (VOCs) in the presence of sunlight to create ozone, which is a major respiratory irritant.
  2. Acid Rain: Like SO₂, NOx also contributes to acid rain by forming nitric acid in the atmosphere.
  3. Human Health: NO₂ itself is a toxic gas that irritates the airways and can aggravate respiratory diseases.
  4. Source: It's formed in *any* high-temperature combustion process (even burning natural gas, which has no sulfur). The high heat causes nitrogen (N₂) from the air to react with oxygen (O₂).

The gold standard and most common method for measuring NOx is Chemiluminescence.

• Principle: This method relies on a specific chemical reaction that produces light.
• How it works:
  1. The Reaction: The analyzer feeds the gas sample into a reaction chamber where it is mixed with a high concentration of ozone (O₃), which is generated by the analyzer itself.
  2. Nitric Oxide (NO) in the sample immediately reacts with the ozone: `NO + O₃ → NO₂* + O₂`
  3. The Light: This reaction creates an "excited" form of NO₂ (NO₂*). This molecule is unstable and immediately decays, releasing its extra energy as a flash of light (photons).
  4. The Measurement: A very sensitive light detector (a photomultiplier tube) measures the intensity of this light. The amount of light produced is directly proportional to the concentration of NO in the sample.
• Measuring Total NOx (NO + NO₂):
  • Since the analyzer only detects NO, a "converter" is needed to measure total NOx.
  • The sample gas is first passed through a "NOx converter" (a heated catalytic chamber) *before* it enters the reaction chamber.
  • This converter uses high heat (e.g., 600-800°C) or a catalyst (e.g., molybdenum) to break down any NO₂ in the sample into NO: `NO₂ + heat → NO + O`
  • Now, the gas stream contains the *original* NO plus the *converted* NO₂. This "total NOx" stream then enters the reaction chamber to be measured.
  • The analyzer can switch between measuring "NO" (bypassing the converter) and "NOx" (passing through the converter) to report both values.

Yes, NDIR (Non-Dispersive Infrared) can also be used to measure NO and sometimes NO₂, though chemiluminescence is generally considered more accurate and sensitive, especially at very low levels.

• How it works: It works on the same principle as for SO₂ or CO. The NDIR analyzer is set to measure the absorption of infrared light at the specific wavelength that NO molecules absorb.
• Gas Filter Correlation (GFC): To improve accuracy and reduce interference from other gases (like water vapor), a more advanced NDIR technique called Gas Filter Correlation is often used. This method is more stable and reliable.
• Pros: Simpler, more robust, and often cheaper than a chemiluminescent analyzer. Doesn't require an ozone generator.
• Cons: Typically less sensitive (higher detection limit) than chemiluminescence. May have more cross-interference from other gases.
• Note: Some multi-gas analyzers use a single NDIR bench to measure CO, SO₂, and NO, making them a cost-effective package.

• What it is: CO stands for Carbon Monoxide. It is a colorless, odorless, and highly toxic gas.
• Why it's monitored:
  1. Incomplete Combustion: This is the *primary* reason from a process standpoint. CO is a product of incomplete combustion. Its presence means the boiler or engine is not burning its fuel efficiently. High CO levels mean fuel is being wasted and soot is likely being produced.
  2. Human Health: CO is extremely dangerous. It displaces oxygen in the bloodstream, leading to asphyxiation. While stack emissions are diluted, high CO is a major indicator of a dangerous process upset.
  3. Regulatory Requirement: Because it indicates poor combustion, CO is often regulated as a "performance" indicator. Many permits have a CO limit to ensure the source is being operated efficiently and cleanly.

The most common method by far for measuring CO in a CEMS is Non-Dispersive Infrared (NDIR).

• Principle: As with SO₂, CO molecules have a unique "fingerprint" in that they absorb infrared (IR) light at a very specific wavelength.
• How it works:
  1. An IR light source shines through a sample cell containing the flue gas.
  2. An optical filter (or Gas Filter Correlation wheel) isolates the specific wavelength of IR light that CO absorbs.
  3. A detector on the other side measures how much of this light was absorbed by the CO molecules.
  4. The amount of absorption is directly proportional to the CO concentration.
• Why NDIR? It is a very robust, reliable, and cost-effective technology. It's well-understood and provides excellent accuracy for the concentration ranges typically seen in flue gas.

Opacity is a measure of the "obscuration" of light by particulate matter (PM) in the flue gas. It is a measurement of how *non-transparent* the exhaust is.

• It is expressed as a percentage (%).
• 0% Opacity: The flue gas is perfectly clear. No light is blocked. (e.g., a natural gas boiler on a clean day).
• 100% Opacity: The flue gas is completely opaque. All light is blocked. (e.g., thick, black smoke).
• Regulatory limits are typically set at low values, such as 20% opacity, averaged over a 6-minute period.
• It is *not* a direct measure of PM *mass* (like mg/m³), but it is a very good, continuous *indicator* of PM emissions. A spike in opacity indicates a problem, such as a failure in a baghouse or electrostatic precipitator.

Opacity is measured using an "in-situ" (in-stack) instrument called a Transmissometer.

• Principle: It measures the transmission of a light beam across the stack.
• Components:
  1. Transmitter (Light Source): Mounted on one side of the stack. It projects a calibrated beam of light (usually a specific wavelength of visible light) across the stack.
  2. Receiver (Detector): Mounted on the *opposite* side of the stack, precisely aligned with the transmitter. It measures the intensity of the light it receives.
  3. Air Purge System: A system of blowers that continuously force clean air over the transmitter and receiver lenses. This is critical to keep the lenses free of soot and dust, which would cause a false high opacity reading.
• How it works:
  • The instrument is first calibrated in a clean stack (0% opacity) to know what 100% light transmission looks like.
  • During operation, particulate matter (soot, ash) in the flue gas scatters and absorbs the light beam.
  • The receiver measures the *loss* of light intensity.
  • Opacity is calculated from this loss. For example, if 15% of the light is blocked by smoke, the opacity is 15%.

This is a correction factor. The opacity monitor (transmissometer) measures opacity at *one specific point* in the stack, across the path length of the monitor (e.g., 20 feet). However, the regulatory limit applies to what a human observer would see *at the stack exit*.

• If the CEMS monitor is located in a part of the stack that has a different diameter than the final stack exit, the reading must be corrected.
• Example: Measuring in a 20-foot wide duct when the stack exit is only 10-feet wide. The light is traveling through *more* smoke in the 20-foot duct, so the opacity reading will be artificially high.
• A "stack exit correlation" or "path length correction" factor is programmed into the DAHS. It's a simple ratio of the monitor path length to the stack exit path length, which scales the reading to be equivalent to the stack exit.

The main challenge in NDIR (infrared) analysis is that different gases can absorb light at similar or overlapping wavelengths. This is called "cross-interference."

• Water Vapor (H₂O): This is the biggest interference for NDIR. Water molecules are strong absorbers in the IR spectrum. This is the primary reason "cold-dry" systems are used—to remove the water and eliminate this interference.
• Carbon Dioxide (CO₂): CO₂ is also a strong IR absorber and can interfere with the CO measurement if their absorption bands overlap.
• How to correct for it:
  1. Optical Filters: High-quality filters are used to select a *very* narrow wavelength band for the target gas (e.g., CO) that is *not* absorbed by the interfering gas (e.g., CO₂).
  2. Gas Filter Correlation (GFC): This is a more advanced method. The analyzer uses a spinning wheel with two cells: one filled with a high concentration of the target gas (e.g., CO) and one filled with a neutral gas (e.g., Nitrogen). This allows the detector to very precisely distinguish the target gas from any background interference.
  3. Dual Wavelength: The analyzer measures at two wavelengths—a "measure" wavelength and a "reference" wavelength—and uses the difference to mathematically cancel out interferences.

The most common method, especially in hot, in-stack probes, is the Zirconium Oxide (ZrO₂) sensor.

• Principle: It's an electrochemical cell. A solid-state sensor made of zirconium oxide (a ceramic) is coated with platinum electrodes.
• How it works:
  1. The sensor is heated to a very high temperature (e.g., 700-800°C).
  2. One side of the sensor (the inner electrode) is exposed to a "reference" gas (usually ambient air, which has 20.9% O₂).
  3. The other side (the outer electrode) is exposed to the flue gas.
  4. At high temperature, the zirconium oxide becomes a solid electrolyte, allowing oxygen *ions* (O²⁻) to pass through it.
  5. The *difference* in oxygen concentration between the reference air and the flue gas creates a *voltage* (a potential) between the two electrodes.
  6. This voltage is measured and, based on the Nernst equation, is converted into a very_precise O₂ concentration percentage.
• Other methods: Paramagnetic analyzers are also used (O₂ is one of the few gases that is paramagnetic, or attracted to a magnetic field), but the ZrO₂ probe is more common for CEMS.

This is a third, distinct type of extractive system that combines in-situ and extractive principles. It's very different from hot-wet or cold-dry.

• How it works:
  1. An in-stack probe extracts a small, constant flow of sample gas.
  2. At the probe tip *inside the stack*, this sample gas is immediately and precisely *diluted* with a large, known-volume of clean, dry "instrument air." A common dilution ratio is 100:1.
  3. This "diluted" sample is now non-corrosive, non-condensing (its dew point is below ambient temperature), and clean.
  4. This clean, diluted sample is then transported through a simple, *unheated* plastic tube (like Teflon) to the analyzers, which can be located far away.
  5. The analyzers measure the very low, diluted concentration (e.g., 1 ppm). The DAHS then multiplies this reading by the known dilution factor (e.g., 100) to get the true stack concentration (e.g., 100 ppm).
• Pros:
  • Extremely reliable. It eliminates the complex sample conditioning system (no chillers, pumps, or heated lines).
  • Analyzers last longer because they only see clean, dry, non-corrosive gas.
  • Very low maintenance.
• Cons:
  • Requires extremely sensitive analyzers (to measure the diluted concentrations).
  • The precision of the dilution orifice is critical.

This is the most common type of "in-situ" analyzer, very similar to an opacity monitor's design.

• How it works: It's a "sender/receiver" or "transmitter/reflector" system.
• A beam of light (e.g., IR or UV) is sent from a transmitter unit on one side of the stack.
• The beam travels *across the entire stack path* and is measured by a receiver unit on the other side (or hits a reflector that bounces it back to a detector on the same side).
• The analyzer measures the absorption of specific light wavelengths *by the entire gas path*.
• This gives a true, averaged reading across the stack's full diameter.
• Example: An "in-situ" (cross-stack) TDL (Tunable Diode Laser) analyzer is often used to measure Ammonia (NH₃) in real-time.

"Quench" or "quenching" is a common source of interference in chemiluminescence analyzers.

• The Problem: The light-producing reaction (`NO + O₃ → NO₂* + O₂`) can be "quenched" or "dampened" by the presence of other molecules, primarily CO₂ and H₂O (water vapor).
• These interfering molecules collide with the "excited" NO₂* molecule and "steal" its energy, causing it to return to a normal state *without* releasing a photon of light.
• The Result: The analyzer "sees" less light and reports a NOx value that is *falsely low*.
• The Solution:
  1. Cold-Dry System: Using a "cold-dry" sample system effectively solves this by removing the water (H₂O), which is a major quenching agent.
  2. Quench Correction: Modern "hot-wet" analyzers that use chemiluminescence have a built-in sensor (e.g., a CO₂ or moisture sensor) and use a correction algorithm in their software to compensate for the calculated quenching effect.

This is a fundamental, automated quality-assurance check that is required by law (e.g., under Part 75). It must be performed at least once every 24-hour "operating day."

• Purpose: To check for "drift"—the slow, natural tendency of an analyzer's reading to move away from its true value over time.
• Process:
  1. The DAHS automatically puts the CEMS into "Calibration Mode" (data is flagged and not used for compliance).
  2. The system injects a "zero" gas (e.g., Nitrogen or "zero air," with 0 ppm of the pollutant) and records the analyzer's reading.
  3. The system then injects a "span" gas (a certified cylinder of gas with a known, high concentration, e.g., 80 ppm of SO₂) and records the analyzer's reading.
• Pass/Fail: The DAHS compares the analyzer's readings to the known values of the gas. The difference is called the "drift."
  • If the drift is *within* the allowed tolerance (e.g., +/- 2.5% of the span value), the CEMS passes the check and returns to normal operation.
  • If the drift *exceeds* the tolerance, the CEMS has "failed" its daily cal. It is considered "out of control," and all data from that point forward is *invalid* until a technician investigates, fixes the problem (e.g., performs a full calibration), and passes another check.

• Calibration Drift Check (Daily "Cal Check"):
  • This is a *passive check*.
  • You inject zero and span gas and *record* the reading.
  • You *do not* adjust the analyzer.
  • Its purpose is only to *verify* that the analyzer is still within its allowed drift tolerance.
• Calibration (or "Adjustment"):
  • This is an *active adjustment*.
  • You inject zero and span gas.
  • If the reading is off, you *adjust* the analyzer's "zero" and "span" settings so that its output *matches* the known value of the gas bottles.
  • This is what a technician does to *fix* an analyzer that has failed its drift check. After the adjustment, another cal check must be run to prove the adjustment was successful.

A Cylinder Gas Audit (CGA) is a quarterly (once every 3 months) Quality Assurance requirement.

• Purpose: It's a "spot check" to ensure the CEMS is accurate over its entire measurement range.
• Process:
  1. The CEMS must be operating normally (it cannot be in calibration or maintenance).
  2. A CEMS technician (or 3rd-party tester) challenges the *entire* CEMS (from the probe) with at least two, and often three, *different* concentrations of "audit" gas.
  3. These "audit" gases must be "NIST-traceable" (certified) and *must not* be the same cylinders used for daily calibrations.
  4. The concentrations are typically:
     • Low-Range: 20-30% of span
     • Mid-Range: 40-60% of span
     • (Sometimes a High-Range is also used)
  5. The CEMS response is recorded for each gas.
• Pass/Fail: The CEMS reading for each gas must be within a certain accuracy, typically +/- 15% of the bottle's value or a small ppm difference (e.g., +/- 5 ppm), whichever is greater. If it fails, corrective action is required.

The RATA is the "Super Bowl" of CEMS testing. It is the most important, complex, and expensive QA test. It is required annually (once per year) and also after the CEMS is first installed.

• Purpose: The RATA *does not* check the CEMS against gas bottles. It checks the CEMS against an *independent, third-party* testing team using the official EPA Reference Methods. Its purpose is to prove that the CEMS data is accurate relative to the "gold standard" EPA test methods.
• Process:
  1. An independent stack testing company sets up their own CEMS equipment (the "Reference Method" or RM) on the stack, using a separate set of test ports.
  2. Both the plant's CEMS and the Reference Method CEMS are run simultaneously for a period of time.
  3. A "test run" consists of comparing the data from the two systems (e.g., over a 21-minute period).
  4. A full RATA consists of at least *nine* (9) successful test runs.
• Pass/Fail:
  • The nine (or more) data pairs are collected and a statistical analysis (using standard deviation and "t-values") is performed.
  • The result is a "Relative Accuracy" (RA) percentage.
  • To pass, the CEMS must have a Relative Accuracy of less than a certain amount, typically 10% (or 7.5% for some programs).
  • If the CEMS *fails* the RATA, it is a major compliance problem. The system is "out of control" and all data is invalid until the CEMS is repaired and can pass a new RATA.

The bias test is part of the RATA result.

• Bias: This is a statistical term that means the CEMS is *consistently* reading higher or lower than the Reference Method.
• Example: The RATA shows that your CEMS *passes* (RA is 8%), but it *consistently* reads 5 ppm lower than the reference method. This is a "low bias."
• The Test: A statistical "t-test" is performed on the RATA data to see if this bias is "statistically significant."
• The Result (Bias Adjustment Factor - BAF):
  • If the bias *is* statistically significant, the CEMS has failed the bias test.
  • The facility must then apply a "Bias Adjustment Factor" (BAF) in the DAHS software.
  • From that point forward, the DAHS must *mathematically correct* all CEMS data to account for the bias. In our example, it would have to *add* 5 ppm to every reading before reporting it.
  • This is undesirable, as it makes your reported emissions higher. The goal is to have no bias.

• Range: This is the maximum *measurement capability* of the analyzer. For example, an analyzer might have a 0-1000 ppm range. This is set in the hardware.
• Span: This is a *regulatory* value. It is the "full scale" value used for calibration and reporting. The span must be set high enough to measure the highest expected emissions, but low enough to be accurate.
  • For example, your permit limit might be 100 ppm. Your analyzer has a 0-1000 ppm range.
  • You might set your "span" value at 150 ppm (1.5 times the limit).
  • Your daily "span" gas would then be 80-100% of this 150 ppm value (e.g., 130 ppm).
  • Your "zero" gas would be 0-20% of this value (e.g., 10 ppm).
  • This ensures your CEMS is most accurate *in the range where you operate*, not at 1000 ppm.

This is an analyzer that can operate accurately at two different "span" settings.

• Why? This is common for sources that have very different emission levels at different times.
• Example: A combustion turbine.
  • "Low Range": When firing natural gas, NOx emissions might be very low (e.g., 0-50 ppm). A "low span" of 50 ppm is used.
  • "High Range": When firing fuel oil (e.g., during a gas curtailment), NOx emissions might be much higher (e.g., 0-200 ppm). The analyzer must automatically switch to a "high span" of 200 ppm.
• The CEMS must perform separate daily calibration checks (and CGAs, and RATAs) for *both* ranges it operates in.

These are the high-accuracy calibration gases required for all CEMS quality assurance activities (Daily Cals, CGAs, RATAs).

• They are *not* just "a bottle of gas." They are specialty gas mixtures that have been certified by the gas manufacturer according to a strict EPA "Protocol."
• The gas bottle comes with a certificate that states its exact concentration (e.g., 100.5 ppm +/- 1%).
• The certification is "NIST-traceable," meaning the manufacturer's own instruments were calibrated against a "Standard Reference Material" (SRM) from the National Institute of Standards and Technology.
• Using non-protocol gases for CEMS calibrations is a major violation and will invalidate all emissions data.

You can't use a "gas" bottle. Opacity monitors are calibrated using Neutral Density Filters.

• What they are: These are calibrated pieces of "smoky" glass or screens that are certified to block a *known amount* of light.
• The Audit: A CEMS technician (or 3rd-party) will manually insert a set of these filters into the monitor's light path.
  • A "Low" filter (e.g., 10% opacity)
  • A "Mid" filter (e.g., 20% opacity)
  • A "High" filter (e.g., 30% opacity)
• Pass/Fail: The monitor's reading must match the filter's certified value within a small tolerance (e.g., +/- 2% opacity). This is a quarterly or semi-annual requirement.
• Daily Check: For its *daily* check, the monitor has internal filters or a "zero" and "span" reflector that it automatically moves into the light path to check for drift.

A "linearity check" is another quarterly QA test, similar to a CGA. Its purpose is to prove the analyzer responds "linearly" across its measurement range.

• What it is: You challenge the analyzer with three certified gas bottles:
  • Low: 20-30% of span
  • Mid: 40-60% of span
  • High: 80-100% of span (this is often the daily span gas)
• The Test: The analyzer's response for each gas is compared to the certified bottle value.
• The Difference: Under Part 75, this test is called a "Linearity Check." Under Part 60, it's called a "Cylinder Gas Audit (CGA)." The procedure and pass/fail criteria are very similar. The key is that you must test *at least* three points (including zero) to prove the analyzer's response is a straight line.

Data validation is the process (usually automated by the DAHS) of determining whether the data collected from the CEMS is "valid" and can be used for compliance reporting, or "invalid" and must be discarded.

Data is considered INVALID during any of the following periods:

• Calibrations: During the daily zero and span checks.
• Maintenance: When a technician has the CEMS in "maintenance mode" to work on it.
• System Malfunctions: If an analyzer fails, a pump stops, or a heater burns out, the DAHS flags all data as invalid until the system is fixed.
• "Out-of-Control" (OOC) Periods:
  • If the CEMS *fails* its daily calibration check.
  • If the CEMS *fails* its quarterly CGA or Linearity.
  • If the CEMS *fails* its annual RATA.
  • All data is invalid from the point of the failed test until the CEMS is repaired and *passes* a new test.

Data substitution (or "missing data substitution") is a regulatory requirement (especially under Part 75) that defines what data you must "fill in" when you have gaps of *invalid* data.

• The Problem: You can't just report "zero" emissions when your CEMS is broken. That would create a loophole where a facility could benefit from having a broken CEMS.
• The Solution: The DAHS must automatically "substitute" the missing data using a "missing data algorithm." This algorithm gets progressively more *punitive* the longer the CEMS is broken.
  • Short Gap (e.g., 1-2 hours): The DAHS might fill in the gap with the *average* of the hours just before and just after the outage.
  • Longer Gap (e.g., 8 hours): The DAHS might have to fill in the gap with the *highest* emission value recorded in the last 30 days.
  • Very Long Gap (e.g., >24 hours): The DAHS might have to fill in the gap with the "Maximum Potential Concentration" (e.g., 1000 ppm)—a very high, punitive value.
• The Purpose: This provides a powerful *financial* incentive to keep the CEMS availability high and to fix any problems *immediately*.

This is a performance metric for the CEMS itself, usually calculated quarterly or annually.

• It is the percentage of "source operating hours" for which the CEMS produced *valid* data.
• Formula: `(Hours of Valid Data / Hours the Source was Operating) * 100%`
• Example: The boiler ran for 1000 hours in a quarter. The CEMS was down for calibrations and one repair for a total of 50 hours. The CEMS availability is `(950 / 1000) * 100 = 95%`.
• Requirement: Most regulations require a CEMS to have a minimum availability, such as 90% or 95% on a quarterly basis.
• If availability drops below this, the facility may be subject to fines and will have to use more punitive missing data substitution.

CEMS are complex and exposed to a harsh environment. The most common failures are almost always in the *sample conditioning system* (for extractive CEMS).

1. Filters: The filters at the probe tip and inside the CEMS cabinet get clogged with particulate and must be replaced regularly (e.g., weekly or monthly). This is the #1 maintenance item.
2. Pumps: The sample pump (which pulls the gas from the stack) and the peristaltic pump (which removes water from the chiller) are moving parts that wear out and fail.
3. Sample Line: The heated sample line can develop "cold spots" if a heater section fails, causing condensation and blockage. Leaks can also develop.
4. Chiller (Cooler): The sample cooler can fail, or it can be overwhelmed by a "slug" of water, leading to moisture getting into the analyzers (which can destroy them).
5. Opacity Monitor Lenses: Despite the air purge, the lenses on the transmissometer will eventually get dirty and need to be manually cleaned.
6. Calibration Gas: The compressed gas cylinders for daily calibrations run out and must be replaced.

This is a critical, automated preventive maintenance function.

• The Problem: The primary filter at the probe tip in the stack gets clogged with fly ash and soot.
• The Solution: The CEMS controller is programmed to periodically (e.g., once every hour) perform a "blowback."
  1. It briefly closes the sample valve.
  2. It opens a valve connected to a high-pressure "instrument air" line.
  3. It shoots a short, powerful blast of high-pressure air *backwards* through the probe and out the tip.
  4. This blast of air blows the accumulated "filter cake" of ash off the filter, cleaning it.
  5. The system then returns to sampling.
• This automated cleaning dramatically extends the life of the probe filter.

"Response time" is the time it takes for the *entire CEMS* to see and respond to a change in emission concentration. It is a required test, often done semi-annually.

• Definition: The time (in minutes) from when a calibration gas is injected at the *probe tip* (to simulate a change in stack gas) until the analyzer's reading reaches 95% of the final, stable value.
• Why test? This checks for "lag" in the system.
  • A long response time (e.g., 10 minutes) could be caused by a partially plugged filter, a weak sample pump, or a leak in the sample line.
  • Regulations require the response time to be under a certain limit (e.F., 15 minutes) to ensure the CEMS is capturing real-time emission events and not "smearing" them out over time.
• This test challenges the *entire* system, including the pump, sample line, and cooler, not just the analyzer itself.

This is a "troubleshooting" question. A logical, step-by-step process is key:

1. Acknowledge the Status: The CEMS is now "Out of Control" (OOC). I must notify plant operations that all emissions data is invalid until the issue is fixed. This starts the "missing data" clock.
2. Verify the Problem:
   • First, I would check the calibration gas bottles. Is the "span" bottle empty or is the pressure too low? Is the regulator valve open? This is a very common and easy-to-fix problem.
   • I would manually re-run the daily calibration check to see if the failure was a one-time glitch.
3. Isolate the Issue:
   • Did it fail the "zero" check or the "span" check? If it failed "span," is it reading high or low? This gives a clue.
   • I would check the system for obvious faults: Are there any alarms on the DAHS? (e.g., "Low Sample Flow," "Chiller Temp High," "Analyzer Fault").
   • I'd walk down the system: Check the sample pump, check the chiller for water, look at the analyzer's internal diagnostics.
4. Perform an "As-Found" Calibration: I would inject the zero and span gases *at the analyzer* (bypassing the sample system) to see if the *analyzer itself* has drifted, or if the problem is in the sample path.
5. Take Corrective Action:
   • If analyzer drifted: I would perform a full "Calibration Adjustment" on the analyzer itself.
   • If analyzer is fine: The problem is in the sample system. The most likely cause is a plugged filter, a leak, or a pump problem. I would change the probe filter and check for leaks.
6. Verify the Fix: After taking corrective action, I *must* run another successful calibration check. Once it passes, the CEMS is back "in control," and I would notify operations that the data is valid again.
7. Documentation: I would document all findings, actions taken, and the "as-found" and "as-left" calibration data in the CEMS logbook, as required by law.

A flow monitor is required for CEMS under programs like Part 75 that require reporting *mass emissions* (e.g., lbs/hr or tons/year).

• The Calculation: To get a mass emission rate, you need two pieces of information:
  1. Pollutant Concentration: `ppm` or `mg/m³` (from the gas analyzer)
  2. Stack Gas Volumetric Flow Rate: `scfh` (Standard Cubic Feet per Hour) (from the flow monitor)
  `Mass Rate (lbs/hr) = Concentration x Flow Rate x Conversion Factors`
• How it's measured:
  • Differential Pressure (Pitot Tube): The most common type. An "averaging" pitot tube (like an S-type) is placed in the stack. It measures the difference between the "impact" pressure of the moving gas and the "static" pressure. This "delta-P" is proportional to the gas velocity (and thus, flow).
  • Ultrasonic: A more modern, high-tech method. Two ultrasonic transducers are mounted across the stack. They send "pings" of sound back and forth. The time it takes the sound to travel *with* the gas flow is faster than the time it takes to travel *against* the flow. This time difference is directly proportional to the gas velocity.

A moisture sensor (or moisture monitoring system) is needed whenever you need to convert emission data from a "dry" basis to a "wet" basis, or vice versa.

• When is it needed?
  1. Cold-Dry CEMS: A "cold-dry" CEMS measures pollutant concentration on a *dry* basis (since the water was removed). However, regulations require reporting emissions on a *wet* basis (as they exist in the stack). A separate moisture sensor (or a default value) is needed to calculate the moisture percentage, which the DAHS then uses to convert the "dry" reading to a "wet" reading.
  2. Mass Emissions: Stack gas flow (from a pitot tube) is measured on a "wet" basis. If you have a "dry" pollutant reading, you must convert one of them to calculate a mass emission rate.
• How it works: In-stack sensors (like TDLs or "wet/dry bulb" systems) can provide a continuous measurement of the % H₂O in the flue gas.

A "time-shared" CEMS is a cost-saving design where a *single set of analyzers* is used to monitor *multiple* stacks or sources.

• How it works:
  1. There is a sample probe and line from each source (e.g., Boiler 1 and Boiler 2).
  2. These lines all go to a "valve manifold" or "stream selector" in the CEMS shelter.
  3. The DAHS controls the manifold, sampling from one source at a time.
  4. For example, it samples Boiler 1 for 15 minutes, then switches to Boiler 2 for 15 minutes, and so on.
• Pros: Much cheaper than buying a dedicated CEMS for each source.
• Cons:
  • Not truly "continuous." You only get a reading from each source for a fraction of the hour.
  • This is only allowed by regulators in specific situations (e.g., for sources that are very similar or have low emissions).
  • The DAHS logic is much more complex, as it must "hold" the last valid reading from Boiler 1 while it's busy measuring Boiler 2.

A TDL analyzer is a highly specific, high-speed analyzer often used for "in-situ" (cross-stack) measurements.

• Principle: It uses a *laser* as its light source.
• How it works:
  1. The laser is "tuned" to emit light at an *extremely* precise and narrow wavelength—one that is *only* absorbed by the target gas molecule (e.g., NH₃, HCl, or CO) and *nothing else*.
  2. This laser beam is shined across the stack.
  3. The detector on the other side measures the absorption at that exact wavelength.
• Pros:
  • Extremely Specific: Because the laser line is so narrow, there is virtually *zero* cross-interference from other gases. This is its biggest advantage.
  • Very Fast: Responds in seconds.
  • In-Situ: Measures the gas directly in the hot, wet stack.
• Common Use: Often used to measure "Ammonia Slip" (NH₃) after an SCR (Selective Catalytic Reduction) NOx control system.

• The Process (SCR): Many sources (like coal plants or gas turbines) use a Selective Catalytic Reduction (SCR) system to control NOx. This system *injects* ammonia (NH₃) into the flue gas. The ammonia reacts with NOx on a catalyst, turning it into harmless Nitrogen (N₂) and water (H₂O).
• The Problem ("Ammonia Slip"): The control system tries to inject the *exact* right amount of ammonia to react with the NOx.
  • If you inject *too little* ammonia, you won't meet your NOx limit.
  • If you inject *too much* ammonia (to be "safe"), the excess, unreacted ammonia "slips" through the catalyst and goes up the stack.
• Why it's bad: "Ammonia slip" is its own form of pollution. It can cause a visible plume and is a regulated gas. It's also very expensive, as you are wasting ammonia.
• The CEMS Connection: To optimize this process, facilities often use *three* CEMS:
  1. A NOx analyzer *before* the SCR (to measure "inlet" NOx).
  2. A NOx analyzer *after* the SCR (to measure "outlet" NOx and prove compliance).
  3. An Ammonia (NH₃) analyzer (often a TDL) *after* the SCR to measure ammonia slip.
  The plant operator uses all three readings to control the ammonia injection rate perfectly.

This is a critical failure, as it's "failing low" and hiding a potential compliance violation. A reading of zero is almost *always* wrong when a source is running.

1. Check Sample System (Most Likely): A zero reading indicates the analyzer is *not getting a sample*.
   • Plugged Filter: The probe filter is completely blocked.
   • Sample Pump Failure: The pump has died.
   • Sample Line Blockage: The line is plugged (e.g., with condensed water that has frozen) or has been crimped.
   • Valve Error: The system is "stuck" in calibration mode, or a "sample" solenoid valve has failed to open.
2. Check the Analyzer:
   • Chemiluminescence (NOx): Check the *ozone generator*. If the O₃ generator has failed, the chemical reaction stops, and the analyzer will read zero. Check the "Ozone" or "Reaction Chamber" fault light on the analyzer.
   • NDIR (if used for NO): Check for an "Analyzer Fault" or "Light Source Failure." If the IR lamp has burned out, it will read zero.
3. Check DAHS: Is the DAHS "polling" the analyzer? It's possible the analyzer is working, but the communication link (e.g., 4-20mA loop) has failed, causing the DAHS to read zero.

The *most* likely cause is a loss of sample flow (pump or plug), followed by a failure of the ozone generator.

This indicates a total loss of the light beam. The cause is either a *real* (and catastrophic) plant event or, more likely, a *monitor failure*.

1. Check Plant Operations:
   • First, I would call the control room. Did the electrostatic precipitator (ESP) or baghouse "trip" (shut down)?
   • If so, the 100% opacity is *real*, and this is a major compliance event.
2. Check Monitor Alignment:
   • If the plant is fine, the monitor has failed.
   • A 100% reading means "no light." The most common cause is that the transmitter and receiver are *no longer aligned*.
   • Vibration from the stack can cause the mounting bolts to loosen, and the transmitter can "droop," so it's no longer pointing at the receiver. I would physically check the alignment.
3. Check Hardware Failure:
   • The *light bulb* in the transmitter has burned out.
   • The *detector* in the receiver has failed.
   • Power has been lost to *one* of the two units.
4. Check for Blockage:
   • Both air purge blowers have failed, and the lenses are *completely* caked in soot. This is less likely to be "sudden" but is possible.
   • A technician may have left a "dust cap" or "shipping cover" on one of the lenses after maintenance.

This is a summary "big picture" question. The answer revolves around the consequences of failure.

• Legal & Financial Consequences:
  • The CEMS is a *legal compliance* instrument. The data it produces is used in legal reports submitted to the government.
  • If the CEMS data is inaccurate, incomplete, or invalid, the facility is in *violation* of its operating permit.
  • This can lead to "Notices of Violation" (NOVs), massive fines (often calculated per *day* of non-compliance), and in extreme cases, forced shutdowns or even criminal charges.
• Operational Consequences:
  • As discussed, the "missing data substitution" rules are highly punitive.
  • Every hour the CEMS is down (due to lack of maintenance) costs the company money, as they must report artificially *high* emissions, which uses up valuable (and expensive) emissions allowances.
  • Poor CEMS performance (e.g., failed a RATA) can result in months of invalid data, leading to a compliance nightmare.
• Environmental & Ethical Consequences:
  • The CEMS is the "watchdog" that ensures the plant's pollution controls (like SCRs and scrubbers) are working.
  • If the CEMS is not maintained, it could "fail low" and hide a major pollution event, allowing the plant to release illegal amounts of SO₂ or NOx, which harms public health and the environment.
  • Proper maintenance and QA/QC ensures the CEMS is providing accurate, reliable, and legally-defensible data, which is its one and only job.