Enlarge this picture Figure 1. Part 75 Calibration Error Test Data

While fossil fuels are an integral resource of energy and a basis for other materials, they have been a blessing and a curse. In 1995 the EPA enacted the Acid Rain Program in an effort to reduce the amount of SO₂ and NO_X emitting from sources like coal-fired electric utilities, petroleum refineries, metal ore plants and cement manufacturing facilities.

In perhaps the EPA's greatest success story, the United States’ comprehensive program has reduced SO₂ and NO_X emissions by nearly 10 million tons per year.^[1] This has significantly reduced the amount of acid rain that once polluted our lakes and streams, and threatened to destroy plant and aquatic life. Today, the Earth’s natural filters (i.e. vegetation) are flourishing in the abundance of greenhouse gases such as CO₂.

The lynchpin of this program is the monitoring processes detailed by the Code of Federal Regulations title 40 CFR Part 75. This legislation provided a market-based platform that established a pollutant emission cap or baseline allowed by each source.^[2]

The primary method of stack sample extraction, analysis and reporting the amount of SO₂, NO_X, CO₂ and oxygen emitted is the continuous emissions monitoring system (CEMS). The accuracy or certification of the CEMS unit is validated by a source’s Part 75 compliant Quality Assurance/Quality Control, or QA/QC, plan. Upon initial certification, the CEMS unit must perform daily calibration error and flow interference tests.

Annual Relative Accuracy Test Audits, or RATA, quarterly linearity and daily calibration tests are dependent upon the stability and accuracy of the analyzer. Whether using absorption or luminescence detection methods, a known pollutant standard, called a “protocol,” is required to create calibration curves for the sample comparison. Protocol cylinders are typically referenced to the gas manufacturer’s intermediate standard (GMIS) or a certified reference material (CRM). Both the GMIS and CRM standards are traceable to the National Institute of Standards and Technology (NIST). With monetary penalties looming and the prospect of negative environmental impact, certain analyzer calibration errors are tightly controlled, as illustrated in Figure 1.^[2]

Calibrating gases

Enlarge this picture Figure 2. Material Permeation Factor (P) x1010 @ 25°C

Construction and design materials can have a variety of effects on a CEMS’ calibration gas system.

The calibration gas system must reduce the cylinder pressure to a safe working pressure while maintaining gas integrity. A pressure regulator performs this task. Gas integrity may be compromised by permeation or leak paths. Figure 2 illustrates the permeation factors of various materials. The driving force of permeation is a function of the difference (delta) in partial pressure across a barrier for a given compound.^[3] Even in controlled shelter environments, moisture and oxygen can permeate through Teflon-lined cylinder hoses or Teflon-wrapped national pipe thread, or NPT, connections, depleting critical SO₂ and NO_X levels.

Exposure to moisture yields an acidic phase, which leads to corrosion and particle generation within the regulator. When the calibration gas’ integrity is compromised, the analyzer’s data acquisition and handling system (DAHS) records the incorrect information. The daily calibration error tolerance can be exceeded, especially when the span value for SO₂ and NO_X is less than 200 ppm.

In an effort to maintain calibration gas integrity, the electric utility or CEMS integrator can incorporate a few simple design features into the calibration gas control system. First, the gas regulator should be mounted on a panel. This allows the technician to hard-pipe stainless tubing directly from the regulator to the analyzer. Panel-mounting prevents movement of the regulator, which can compromise the compression fitting to tube seal. In some cases a diffusion-resistant valve is installed to isolate the tubing from the regulator. Operators should resist the urge to use lower-cost needle valves, which at best only provide 1 x 10^-5 helium cc/sec leak integrity between the packing and the rotating stem. The optimum choice is a 316L stainless diaphragm valve that maintains a 1 x 10^-9 helium cc/sec metal-to-metal seal.

The regulator

On the left side (3A) is an example of tapped threads. This can result in cracks that form potential leak paths. On the right side (3B) is a threaded joint manufactured using single point machining methods.

Typically, a panel-mounted regulator will be upstream of the isolation valve. Three key areas of regulator design that impact diffusion are the NPT inlet/outlet ports, internal surface finishes and particulate filtration.

In an effort to control costs, some manufacturers use a tapping process to form the port threads. The tap must stop at the bottom of the thread and reverse direction to exit the port. This process leaves four full-length perpendicular lines on the thread. Each line creates a potential leak path, as illustrated in Figure 3A. Tapping leaves a rough-thread face finish, which leads to galling or seizing of stainless connections. Alternate methods such as single-point machining exist, but it takes twice as long to complete the thread. Leading manufacturers have developed proprietary thread machining processes that yield excellent thread face finishes without tooling marks, as illustrated in Figure 3B. The thread surface finish is critical in maintaining a 1x10^-9 helium cc/sec regulator leak rate.

Preserving the finish

The filter encapsulating the regulator seat prevents particulate from causing leaks across the seat.

It is necessary to maintain high-quality surface finishes on whetted areas and the diaphragm seal. As the machined surface becomes rougher, the peaks and valleys become pronounced, providing entrapment sites for moisture and particulates. It is in these crevices that SO₂ and NO_X react to form an acidic phase, leaving the machined peaks susceptible to localized corrosion.

Most manufacturers seal the stainless diaphragm in one of two ways. The low-cost method utilizes a recessed Teflon O-ring. This method enables the manufacturer to use zinc die-cast bonnets and rougher surface finishes. A negative of this design is the O-ring’s high surface area, which when exposed to atmosphere can allow greater permeation. Also, only modest leak integrity (1 x 10^-8 helium cc/sec) is achieved. Alternatively, a metal-to-metal diaphragm seal makes a 1 x 10^-9 helium cc/sec leak rate sustainable by eliminating the Teflon O-ring as a permeation source. The metal-to-metal seal is only possible with high-quality surface finishes and barstock bonnets.

Analyzer reliability can be affected by particulate generation. This is especially true of extractive systems that require filters capable of removing particles larger than 0.5 µm. Most regulator manufacturers incorporate a 10- to 20-micron inlet filter to protect the regulator seat from external debris. One micron is equivalent to 39.37 µ inches. However, most particles are generated during assembly of peripheral components and internal seat parts. For this reason internal filters that encapsulate the first and second stage seats of a dual-stage regulator provide the highest filtering efficiency, as illustrated in Figure 4. A byproduct of this filtration is long-term reliability, which is a requirement for automated CEMS.

Ongoing efforts to monitor and limit the emission of greenhouse gases like SO₂ and NO_X will be driven by the world population’s insatiable demand for energy. To this end, a properly designed calibration gas system continues to be the benchmark for future EPA programs. PE