Laboratory investigation of three distinct emissions monitors for hydrochloric acid

ABSTRACT

The measurement of hydrochloric acid (HCl) on a continuous basis in coal-fired plants is expected to become more important if HCl standards become implemented as part of the Federal Mercury and Air Toxics Standards (MATS) standards that are under consideration. For this study, the operational performance of three methods/instruments, including tunable diode laser absorption spectroscopy (TDLAS), cavity ring down spectroscopy (CRDS), and Fourier transform infrared (FTIR) spectroscopy, were evaluated over a range of real-world operating environments. Evaluations were done over an HCl concentration range of 0–25 ppmv and temperatures of 25, 100, and 185 °C. The average differences with respect to temperature were 3.0% for the TDL for values over 2.0 ppmv and 6.9% of all concentrations, 3.3% for the CRDS, and 4.5% for the FTIR. Interference tests for H₂O, SO₂, and CO, CO₂, and NO for a range of concentrations typical of flue gases from coal-fired power plants did not show any strong interferences. The possible exception was an interference from H₂O with the FTIR. The instrument average precision over the entire range was 4.4% for the TDL with better precision seen for concentrations levels of 2.0 ppmv and above, 2.5% for the CRDS, and 3.5% for the FTIR. The minimum detection limits were all on the order of 0.25 ppmv, or less, utilizing the TDL values with a 5-m path. Zero drift was found to be 1.48% for the TDL, 0.88% for the CRDS, and 1.28% for the FTIR.

Implications: This study provides an evaluation of the operational performance of three methods/instruments, including TDL absorption spectroscopy (TDLAS), cavity ring down spectroscopy (CRDS), and FTIR spectroscopy, for the measurement of hydrochloric acid (HCl) over a range of real-world operating environments. The results showed good instrument accuracy as a function of temperature and no strong interferences for flue gases typical to coal-fired power plants. The results show that these instruments would be viable for the measurement of HCl in coal-fired plants if HCl standards become implemented as part of the Federal Mercury and Air Toxics Standards (MATS) standards that are under consideration.

Introduction

In the United States, consent decrees, state laws, and the proposed Federal Mercury and Air Toxics Standards (MATS) standards for coal-fired plants have driven the need for robust (U.S. Environmental Protection Agency [EPA], 2016), accurate, and certifiable continuous emissions monitors (CEMS) for mercury, particulate matter (PM), ammonia (NH₃), and acid gases. One of the key elements of successful control programs is the development of instrumentation that is simple to use, robust, and accurate under a range of different operating environments. Hydrochloric acid (HCl) is another gas that can readily be measured with tunable diode laser absorption spectroscopy (TDLAS). Under the current rules, HCl is not continuously measured at most plants. However, if HCl standards eventually become implemented as part of the MATS standards, it is likely that many plants will elect to continuously monitor HCl. The typical level of HCl expected at plants with highly controlled stack emissions is expected to be much less than 2 ppm. With such low emission levels, it is important to understand the detection limit and any potential measurement interferences for these instruments. As this application has not been implemented, experience with instruments for this purpose is limited. Also, the sensitivity and reproducibility requirements for the HCl measurements that might be needed may be more stringent, as unlike other measurements such as the NH₃ measurements that are used for process control, the acid gas measurements are for CEMS applications and ultimately would need to meet EPA requirements.

As the measurement requirements in power plants have expanded over the last decade, this has led to the development of more advanced instruments for power plant applications. The application of tunable diode lasers (TDLs) for the measurement of a wide range of gases has grown considerably over this time (Himes et al., 2006; Von Drasek et al., 2006). This includes NH₃, HCl, carbon monoxide (CO), carbon dioxide (CO₂), hydrofluoric acid (HF), oxygen (O₂), ethane (C₂H₆), methane (CH₄), and water (H₂O) (Lackner, 2007). TDLAS provides advantages in that the laser can be tuned to a wavelength that is very specific to a particular gas of interest (Ahlber et al., 1994; Schiff et al., 1994; Fried et al., 1991). TDLAS spectroscopy is now one of the main techniques utilized for monitoring and controlling NH₃ emissions from power plants and other stationary source generation applications (Himes et al., 2006). For the NH₃ application, TDLAS has undergone extensive testing for power plant applications (Himes and Pisano, 2009; Himes et al., 2012). This includes both laboratory testing, as well as field testing and installation.

Continuous-wave Cavity ring down spectroscopy (CW-CRDS) is another technique that is being developed for power plant and other applications. Based on absorption spectroscopy, CRDS works by attuning light rays to the unique molecular fingerprint of the sample gas. By measuring the time it takes the light to fade or “ring down” in a sample cell or cavity, an accurate molecular count can be obtained in milliseconds. The time of light decay, in essence, provides an exact, noninvasive, and rapid means to detect contaminants in air and other gas matrices. As the principle of operation is based on the measurement of a decay rate rather than an absolute absorbance, this is one reason for the increased sensitivity over traditional absorption spectroscopy, as the technique can achieve very long path lengths. Recent developments have made the commercialization of this technique possible. It was shown that compact, relatively inexpensive, and widely available CW tunable diode lasers (TDLs) can substitute for the costly, cumbersome pulsed lasers previously used in CRDS-based research, allowing development of affordable and practical instruments for commercial use.

Fourier transform infrared (FTIR) spectrometers are used to make spectroscopic measurements in many diverse disciplines, and FTIR technology is being used in the development of many commercial instruments. Most of the commercial instruments employed in this type of infrared spectroscopy use an interferometer originally designed by Michelson to measure the speed of light. The interferometer works by taking the entire incident beam of radiation from the source and dividing it into two paths with a beam splitter. One of the paths goes to a fixed mirror, while the other path goes to a moving (translating) mirror. When the position of the translating mirror is continuously varied along an axis collinear to the source, an interference pattern is generated as the two phase-shifted beams interfere with each other. A Fourier transformation is then used to convert the raw data into an actual spectrum, aided by the development of fast computers. To obtain the best analysis of measured spectra, the region used for analysis needs to be chosen to maximize the influence of the gas of interest and to minimize the influence of other gases. As HCl absorbs in a region relatively free of interfering gases, FTIR for the measurement of HCl is another method that has been developed.

The University of California at Riverside (UCR), in conjunction with the Electric Power Research Institute (EPRI), has constructed a laboratory facility for simulating in situ measurements for instrument evaluations (Himes et al., 2006, 2009, 2012; Himes and Pisano, 2007a, 2007b, 2007c, 2009; Dene et al., 2011). The role of the laboratory is to verify the response of various instrumentation over temperature ranges and moisture levels that are representative in typical flue gases at coal-fired power plants. For the laboratory experiments, gas mixtures with temperatures between 300 and 400 °C and up to 12% moisture are typically utilized, which represents flue gas conditions in a power plant exhaust stream found past the economizer but prior to the air heater. The laboratory apparatus was constructed and finalized in 2007 and has been used to successfully evaluate eight different TDL instruments for NH₃ from six manufacturers (Himes et al., 2006, 2009; Himes and Pisano, 2007a, 2007b, 2007c; Dene et al., 2011).

The focus of this current work is on the evaluation and verification of the operational performance and specifications several HCl instruments over a range of real-world operating environments (Dene et al., 2013). In this study, three methods/instruments were evaluated, including TDL spectroscopy, cavity ring down spectroscopy, and FTIR spectroscopy. This effort builds on a previous study, which focused more on evaluating TDL spectroscopy for HCl measurements (Dene et al., 2011). In this study, the instruments were evaluated for verification of detection limits and linearity, and testing for potential interfering gases. The testing was conducted at the UCR’s spectroscopy evaluation laboratory. Evaluations were done over an HCl concentration range of 0–25 ppmv. The evaluations were conducted at temperatures of 25, 100, and 185 °C. Tests of potential interference gases were also conducted at 185 °C for 12% H₂O, SO₂, and a matrix of CO, CO₂, and NO. This work also complements recent field tests that were conducted to evaluate the performance of various HCl instruments at a power plant equipped with an electrostatic precipitator with low-sulfur sub-bituminous coal (Martin et al., 2014).

Test methods

Instruments

Three different instruments were evaluated in this study, including a TDL, a CRDS, and an FTIR. The TDL was a Unisearch Associates Inc. (Concord, ON, Canada) LasIR 410 S-Series (Mackay et al., 2010). This instrument was chosen because it is already being used at one coal-fired power plant site for continuous HCl measurements. The CRDS was a Tiger Optics (Warrington, PA) standard-range Tiger-i 2000 HCl unit. This instrument provides very low detection limits and a fast speed of response. The FTIR was a IMACC (Industrial Monitor and Controls Corporation, Red Rock) FTIR that had HCl methods developed for measuring at 25, 100, and 185 °C. A further description of the TDL is provided in EPRI report 1022083 (Dene et al., 2011), and further descriptions of the TDL, CDRS, and FTIR are provided in EPRI report 3002002463 (Dene et al., 2013).

Experimental setup

The sample cell is described briefly here and in greater detail elsewhere (Dene et al., 2013). Briefly, the sample cell is 1.0 m long with an inner diameter (ID) of 5.0 cm. It is made of 314 electropolished stainless steel and is quartz lined. The sample cell has a radiative heater, is insulated, and is capable of a temperature range of 27–427 °C. This sample cell is typical of those commonly used to simulate in situ flue gas TDL measurements for HCl. It is designed to accommodate varying flows of a source gas and dilution air to provide different gas concentrations. The sample cell has a path length of 1.00 m for single pass optics, or 2.00 m for dual pass optics. For the TDL, HCl can be detected in units of ppmv-m when the concentrations are multiplied by the 1 m path length. For this test, the sample cell was modified so that the output from the sample cell served as the inlet to both the cavity ring down spectrometer and the FTIR. The sample cell is maintained at atmospheric pressure.

Calibration and gas and water delivery systems

The calibration system provides precise concentrations of the HCl target gas by utilizing precision gas dilution. The TECO (Franklin, MA) 146C calibrator employs a central processor unit (CPU) and controlled precision mass flow controllers (MFCs). A Linde (Murray Hill, NJ) certified gas cylinder, containing 100 ppmv of HCl gas in an inert (nitrogen) carrier, provided the source gas through a precision, low-flow MFC. The source gas is mixed with zero air in predetermined flow rates using MFCs to achieve the desired concentrations. After the HCl is mixed with dilution air, the effluent is heated to the desired temperature. The HCl concentration is calculated from the cylinder concentration, the span gas flow rate, and the dilution air flow rate and is checked using a Unisearch (Concord, ON, Canada) LasIR Two-Tone FM (TTFM) TDL as a reference.

Gaseous water is introduced into the blended HCl sample gas stream to provide the desired water content. The water generation system consists of a water reservoir, a graduated burette, a positive displacement pump, and a heating system. The water is drawn through the pump into a 1/8-inch stainless steel line, which is routed through a heated aluminum block with a particulate filter to gasify the water, and then enters the diluted HCl gas stream.

Test matrix

The source gas levels for the evaluation were chosen to cover HCl levels that can be found over a wide range of process conditions, which was determined to be in the range of 0.5–5 ppmv based on field tests. Since most exhaust stacks are around 5 m in diameter, and a cross-duct optical setup is required for in situ monitoring, the levels chosen were from 2.5 to 25 ppmv-m, as TDL sensitivity is nominally reported in ppmv-m, where the levels are multiplied by the path length. Since the laboratory sampling cell is only 1 m long, the values chosen for the evaluation were 0, 0.5, 1, 2, 2.5, 5, 10, and 25 ppmv, which yielded levels of 0, 0.5, 1, 2, 2.5, 5, 10, and 25 ppmv-m, similar to the levels found in coal-fired power plant flue gases.

Tests were conducted at three separate temperatures, 25, 100, and 185 °C, with a 0% moisture content. For the 185 °C temperature, testing was done at two moisture levels, 0% and 12%. The 12% moisture level was selected because this was the highest moisture level that is typically seen in power plant combustion. Initial tests were done at an ambient temperature of 25 °C with no moisture as a control only. Each test was conducted over a 15-min interval. The instruments were configured to provide 15 separate 1-min measurements. The 15-min intervals were then repeated five times to obtain the averages and standard deviations provided in the tables and figures. Note that the test matrix was configured to complete all the tests at a specific temperature, varying the concentration of the target gas at each of the moistures investigated.

Additional tests were also run with other gases present in the exhaust stream that could be potential interferences or gases that could bias the TDL measurements. The possible interfering gases were identified from typical known constituents of gases in coal-fired power plant exhaust streams. These were H₂O, CO, CO₂, NO, and SO₂. Also, with the prevalence of selective catalytic reduction (SCR) and/or selective noncatalytic reduction (SNCR) technologies being employed for NO_x control, possible interference from NH₃ was also evaluated.

Results

Standard and temperature corrections

Measuring over a range of different temperatures is important, since spectroscopic instruments must compensate for temperature changes in order to maintain accuracy over a range of temperatures. This is due in part to the fact that all spectroscopic instruments measure a mixing ratio, which is a ratio of the number of molecules of the measured gas to the total number of molecules in a specific volume. As temperature changes, the number of total molecules for a given volume changes. For example, the number of total molecules at 25 °C in 1 L of sample gas is different than the number of total molecules at 200 °C in 1 L of sample gas. An ideal gas law correction is needed to compensate the sample back to the condition of the instrument calibration. Additionally, the ability for any unsymmetrical polyatomic molecule to absorb energy changes with temperature, so the temperature correction algorithm must also deal with how the molecular absorption of the target molecule behaves with changes in temperature. This effect is called the line strength ratio.

The agreement of the instrument readings with the HCl gas standards was assessed by taking 15 1-min averaged samples at each of the test matrix configurations. The results for the measurements for the three instruments at each of the temperature settings are provided in Table 1 for the 0% moisture tests. The values and associated standard deviations are for the 15 1-min averages of the sample gas measurement. The average percent differences at each test point are provided in Table 2 for each instrument for the 0% moisture tests. These results are shown graphically in Figure 1, Figure 2, and Figure 3, respectively, for the TDL, CRDS, and FTIR.

Table 1. Results of for the three instruments for 1.0 meter sampling system at 25 °C, 100 °C, and 185 °C, all with 0% water. The uncertainties represent one standard deviation of the measurement average.

	Baseline 25 °C	100 °C	185 °C
Standard	0% Moisture	0% Moisture	0% Moisture
HCl (ppmv-m)	HCl (ppmv-m)	HCl (ppmv-m)	HCl (ppmv-m)
TDL
0.0	0.04 ± 0.04	0.04 ± 0.05	0.05 ± 0.02
0.5	0.61 ± 0.05	0.59 ± 0.05	0.54 ± 0.03
1.0	1.13 ± 0.11	1.12 ± 0.11	1.11 ± 0.04
2.0	2.16 ± 0.10	2.17 ± 0.10	2.19 ± 0.04
2.5	2.55 ± 0.12	2.59 ± 0.12	2.67 ± 0.18
5.0	4.81 ± 0.19	4.85 ± 0.19	4.75 ± 0.19
10.0	10.16 ± 0.26	10.16 ± 0.26	10.19 ± 0.18
25.0	25.34 ± 0.38	25.71 ± 0.38	25.69 ± 0.22
CRDS
0.0	0.02 ± 0.02	0.02 ± 0.03	0.04 ± 0.03
0.5	0.51 ± 0.03	0.53 ± 0.03	0.51 ± 0.03
1.0	0.99 ± 0.04	1.06 ± 0.04	1.07 ± 0.04
2.0	2.09 ± 0.04	2.11 ± 0.04	2.09 ± 0.04
2.5	2.52 ± 0.18	2.51 ± 0.18	2.48 ± 0.18
5.0	4.65 ± 0.19	4.66 ± 0.19	4.68 ± 0.19
10.0	9.87 ± 0.18	9.77 ± 0.18	9.81 ± 0.18
25.0	24.86 ± 0.22	24.78 ± 0.22	24.88 ± 0.22
FTIR
0.0	0.04 ± 0.04	0.03 ± 0.05	0.05 ± 0.02
0.5	0.53 ± 0.05	0.55 ± 0.05	0.53 ± 0.03
1.0	0.97 ± 0.09	1.13 ± 0.09	1.05 ± 0.04
2.0	2.12 ± 0.09	2.14 ± 0.09	2.14 ± 0.04
2.5	2.54 ± 0.08	2.54 ± 0.08	2.53 ± 0.18
5.0	4.89 ± 0.23	4.83 ± 0.23	4.81 ± 0.19
10.0	10.23 ± 0.24	10.13 ± 0.24	10.22 ± 0.18
25.0	25.77 ± 0.45	26.1 ± 0.45	25.98 ± 0.22

Table 2. Average percent difference of the measured results obtained with respect to the standard gas values obtained with the following tests: 25 °C, 100 °C, 185 °C, and for all tests conducted at 0% moisture.

Cal Gas (25 °C)	Percentage Differences
	25 °C	100 °C	185 °C	All Tests
TDL
0.5	22.0%	18.0%	8.0%	16.0%
1.0	13.0%	12.0%	11.0%	12.0%
2.0	8.0%	8.5%	9.5%	8.7%
2.5	2.2%	3.6%	6.8%	4.2%
5.0	3.8%	3.0%	5.0%	3.9%
10.0	1.6%	1.6%	1.9%	1.7%
25.0	1.4%	2.8%	2.8%	2.3%
Average	7.4%	7.1%	6.4%	7.0%
Average >2.0	2.2%	2.8%	4.1%	3.0%
CRDS
0.5	2.0%	6.0%	2.0%	3.3%
1.0	1.4%	6.0%	7.0%	4.8%
2.0	4.5%	5.5%	4.5%	4.8%
2.5	0.8%	0.4%	0.8%	0.7%
5.0	7.0%	6.8%	6.4%	6.7%
10.0	1.3%	2.3%	1.9%	1.8%
25.0	0.6%	0.9%	0.5%	0.6%
Average	2.5%	4.0%	3.3%	3.3%
Average >2.0	3.0%	3.3%	2.9%	3.1%
FTIR
0.5	6.0%	10.0%	6.0%	7.3%
1.0	3.0%	13.0%	5.0%	7.0%
2.0	6.0%	7.0%	7.0%	6.7%
2.5	1.7%	1.6%	1.2%	1.5%
5.0	2.2%	3.4%	3.8%	3.1%
10.0	2.3%	1.3%	2.2%	1.9%
25.0	3.1%	4.4%	3.9%	3.8%
Average	3.5%	5.8%	4.2%	4.5%

Figure 1. TDL system response over the range of temperatures examined for this study.

Figure 2. CRDS system response over the range of temperatures examined for this study.

Figure 3. FTIR system response over the range of temperatures examined for this study.

For temperature effects, the average percent difference for each matrix point was determined using the following equation (eq 1):

(1)

where d(t)j refers to the average percent difference with respect to temperature between the expected HCl concentration (x_i) from the calibrated standard source and the 15-min average HCl reading (a_i) during the sampling period (corrected for the initial zero air background), and n corresponds to the number of samples for different temperatures at the same moisture. The number of samples for this study was 21, which corresponds to seven concentrations for zero moisture at each of the three temperatures, 25 °C (77 °F), 100 °C (212 °F), and 185 °C (365 °F).

The average differences with respect to temperature were 7.0% for the TDL, 3.3% for the CRDS, and 4.5% for the FTIR. The TDL was characterized separately for concentrations above 2.0 ppmv, since the 1 m path length used for these tests is less than the 5 m of more path lengths that are used at most power plant installations. If only the values above 2.0 ppmv are used for the TDL, the average difference with respect to temperature drops to 3.0%. This is consistent with the results of a previous study, which showed average differences of less than 3% for testing started at 2.5 ppmv (Dene et al., 2011).

The results for the individual test matrix points are within the typical precision limits of the system itself, which is with approximately 3–5%, for many of the test points. In most cases, the percentage differences were below 10%, with the exception of a few points mostly at the lower concentrations. The larger percentage differences could be due to slight differences in the delivery system for specific tests, but do not appear to exhibit distinct trends with respect to either temperature or moisture content. For the TDL, the larger percentage differences at the low concentrations could also be due to the shorter 1 m path length, as opposed to the 5 m used in typical applications. It is significant that the TDL performed as well as it did with such a short path length at the lower concentrations. Since the 0.5 and 1 ppmv-m concentrations approach the 0.3 ppmv-m minimum detection limit, the results are perhaps even better than expected for the TDL.

Interference tests

Laboratory tests were also conducted using gas cylinders with interfering gases that had concentrations similar to those found in coal-fired power plant flue gases. This include a test with a 12% water interferent, an SO₂ interferent, and a matrix of CO, CO₂, and NO interferents. To simplify matters, the cylinders with the interfering gases for the SO₂ and the CO, CO₂, and NO matrix were used as the carrier gases for the HCl into the sample cell instead of zero air, which was used for the earlier tests. The concentrations for the interferences were 2050 ppmv for SO₂ and 1220 ppmv CO, 11.9% CO₂, and 1510 ppmv NO for that test matrix of gases. The results for the measurements for the three instruments for the baseline 0% moisture test and each of the interference tests at 185 °C are provided in Table 3. The average percent differences at each of these test points are provided in Table 4 for each instrument. These results are shown graphically in Figure 4, Figure 5, and Figure 6, respectively, for the TDL, CRDS, and FTIR. The interference tests were conducted only once at each level. It should be noted that additional testing is underway for two different TDLs to evaluate the full range of interferents included in the performance specification-18 (EPA, 2014), including SO₂, CO₂, NO, NO₂, HCHO, O₂, CO, CH₄, and H₂O. These results will be presented in a subsequent paper. Particulates are another potential interferent that is addressed differently for each instrument. The CRDS and FTIR are both extractive methods, so particulates are filtered out of the gas stream before the measurements are made. The TDL, on the other hand, is an in situ method that accounts for the particulate background by scanning areas of the spectral range where there is no HCl absorption to determine the background impact (Dene et al., 2011).

Table 3. Interfering gas test results of the three instruments for 1.0 meter sampling for 0% moisture, 12% moisture, 2050 ppmV SO₂, and 1510 ppmV NO, 1220 ppmV CO, 11.9% CO₂ at 185 °C. The uncertainties represent one standard deviation of the measurement average.

	Baseline, 0% Moisture	12% Moisture	Interference SO2	Interference NO, CO, CO2
Standard	185 °C	185 °C	185 °C	185 °C
HCl (ppmv-m)	HCl (ppmv-m)	HCl (ppmv-m)	HCl (ppmv-m)	HCl (ppmv-m)
TDL
0.0	0.05 ± 0.02	0.06 ± 0.04	0.03 ± 0.04	0.04 ± 0.04
0.5	0.54 ± 0.03	0.57 ± 0.04	0.56 ± 0.05	0.56 ± 0.04
1.0	1.11 ± 0.04	1.18 ± 0.10	1.10 ± 0.11	1.12 ± 0.10
2.0	2.19 ± 0.04	2.13 ± 0.06	2.10 ± 0.08	2.10 ± 0.08
2.5	2.67 ± 0.18	2.59 ± 0.15	2.55 ± 0.14	2.56 ± 0.14
5.0	4.75 ± 0.19	4.77 ± 0.18	4.86 ± 0.19	4.84 ± 0.18
10.0	10.19 ± 0.18	10.23 ± 0.21	10.13 ± 0.24	10.15 ± 0.23
25.0	25.69 ± 0.22	25.77 ± 0.23	25.69 ± 0.31	25.51 ± 0.29
CRDS
0.0	0.04 ± 0.03	0.04 ± 0.04	0.01 ± 0.03	0.01 ± 0.03
0.5	0.51 ± 0.03	0.53 ± 0.03	0.52 ± 0.03	0.53 ± 0.03
1.0	1.07 ± 0.04	1.05 ± 0.04	1.02 ± 0.04	1.02 ± 0.04
2.0	2.09 ± 0.04	2.12 ± 0.04	2.07 ± 0.04	2.08 ± 0.04
2.5	2.48 ± 0.18	2.45 ± 0.18	2.51 ± 0.18	2.51 ± 0.18
5.0	4.68 ± 0.19	4.71 ± 0.19	4.77 ± 0.19	4.75 ± 0.19
10.0	9.81 ± 0.18	9.79 ± 0.18	9.88 ± 0.18	9.86 ± 0.18
25.0	24.88 ± 0.22	24.77 ± 0.22	24.88 ± 0.22	24.86 ± 0.22
FTIR
0.0	0.05 ± 0.02	0.07 ± 0.04	0.04 ± 0.03	0.05 ± 0.03
0.5	0.53 ± 0.03	0.55 ± 0.05	0.53 ± 0.04	0.53 ± 0.04
1.0	1.05 ± 0.04	1.05 ± 0.08	1.03 ± 0.06	1.04 ± 0.06
2.0	2.14 ± 0.04	2.21 ± 0.08	2.12 ± 0.06	2.14 ± 0.06
2.5	2.53 ± 0.18	2.61 ± 0.11	2.55 ± 0.14	2.56 ± 0.13
5.0	4.81 ± 0.19	4.89 ± 0.22	4.90 ± 0.21	4.90 ± 0.21
10.0	10.22 ± 0.18	10.31 ± 0.23	10.18 ± 0.20	10.20 ± 0.21
25.0	25.98 ± 0.22	25.88 ± 0.39	25.62 ± 0.31	25.67 ± 0.33

Table 4. Average percent difference of the measured results obtained with respect to the standard gas values for 0% moisture, 12% moisture, SO₂, and NO, CO, CO₂ all at 185 °C. The uncertainties represent one standard deviation of the measurement average.

Standard HCl (ppmv-m)	Percentage Difference H2O Test	Percentage Difference SO2 Test	Percentage Difference NO, CO, CO2 Test	Percentage Difference Average of Interference Tests
TDL
0.5	14.0%	12.0%	12.4%	12.8%
1.0	18.0%	10.3%	11.9%	13.4%
2.0	6.5%	4.8%	5.5%	5.2%
2.5	3.6%	1.9%	2.3%	2.6%
5.0	4.6%	2.8%	3.2%	3.5%
10.0	2.3%	1.3%	1.5%	1.7%
25.0	3.1%	2.8%	2.1%	2.6%
Average	7.4%	5.1%	5.5%	6.0%
Average >2.0	3.4%	2.2%	2.2%	2.6%
CRDS
0.5	10.0%	4.0%	5.2%	6.4%
1.0	6.0%	1.5%	2.4%	3.3%
2.0	5.5%	3.3%	3.8%	4.2%
2.5	0.5%	0.4%	0.4%	0.4%
5.0	6.8%	4.6%	5.0%	5.5%
10.0	2.3%	1.2%	1.4%	1.6%
25.0	0.9%	0.5%	0.6%	0.6%
Average	4.6%	2.2%	2.7%	3.3%
Average >2.0	3.3%	2.1%	2.3%	2.6%
FTIR
0.5	10.0%	5.3%	6.3%	7.2%
1.0	5.0%	3.3%	3.7%	4.0%
2.0	10.5%	5.8%	6.8%	7.7%
2.5	4.5%	1.9%	2.4%	2.9%
5.0	2.2%	2.0%	2.0%	2.1%
10.0	3.1%	1.8%	2.0%	2.3%
25.0	3.5%	2.5%	2.7%	2.9%
Average	5.5%	3.2%	3.7%	4.2%
Average >2.0	3.3%	2.1%	2.3%	2.6%

Figure 4. TDL system response over the range of interferents examined for this study.

Figure 5. CRDS system response over the range of interferents examined for this study.

Figure 6. FTIR system response over the range of interferents examined for this study.

It is clear from these tests that these potential interferent gasses at concentrations typical to flue gasses in coal-fired power plants do not interfere with the TDL, CRDS, and FTIR instruments, as there was no observed effect in the system response, with the possible exception of H₂O for the FTIR. For the TDL, the results show higher percentage differences for the 0.5, 1.0, and 2.0 ppmv levels, as discussed above, but percentage differences of less than 3% for the testing started at 2.5 ppmv. For the CRDS, the tests results at 2.5 ppmv are very close for all three interference tests. At that level without dilution, the CRDS appears to be very accurate, as it is just over half way to its saturation point. Overall, the tests do not seem to be biased by going through the dilution system, as average percent difference values were 3.2% direct and 2.6% through the dilution system.

The FTIR tests results did show some higher percent differences of 10% and 10.5% (especially at the lower HCl levels) during the H₂O test. Since the FTIR is a broadband IR technique, it is always difficult to totally mitigate effects from H₂O; however, at the higher levels of HCl, the instrument measured through the dilution system and percent differences of the measurements were low (2.2–3.5%). The average percentage difference of 5.5% over all concentrations for the H₂O tests was also higher than the average percentage differences for the other interference tests with SO₂ and the CO, CO₂, and NO matrix.

Although there were no H₂O effects at 12% over a 1-m path for the TDL, higher concentration tests should be considered for the TDL HCl measurements. Measuring at 12% moisture over a 1-m path is not truly representative of what is done at power plants where the levels are around 12% but the path lengths are closer to 5–6 m, so eventually tests should be done at elevated levels to approximate 60–72%-m (path length times concentration).

For moisture effects, the average percent difference for each matrix point was determined by the following equation (eq 2):

(2)

where d(m)j refers to the average percent difference with respect to moisture between the expected HCl concentration (x_i) from the calibrated standard source and the 15-min average SM410 HCl reading (a_i) during the sampling period (corrected for the initial zero air background), and n corresponds to the number of samples of different moistures at the same temperature. For these tests, n is 14, since measurements were made at 185 °C at 0% and 12% moisture levels for each of seven concentrations. The average differences with respect to temperature were 7.2% for the TDL as a whole and 2.8% for values above 2.0 ppmv for the TDL, 3.9% for the CDRS, and 4.5% for the FTIR.

Linearity, precision, and minimum detection limits

Linearity was assessed by a linear regression analysis of the 15-min averaged values for all six tests that were done at the same concentration over the range of moisture, temperature, and interference variables. The six tests include one test at 25 °C, one test at 100 °C at 0% moisture, one test at 185 °C at 0% moisture, one test at 185 °C at 12% moisture, and the two interference tests for SO₂ and the matrix of CO, CO₂, and NO at 185 °C and 0% moisture. Linearity is expressed in terms of slope, intercept, and coefficient of determination (R²) and is shown in Figure 7. All the instruments show good linearity over the range of concentrations evaluated with slopes of 1.022, 0.991, and 1.033, respectively, for the TDL, CRDS, and FTIR, and an R² of 0.998 for all three instruments.

Figure 7. Linear regression of the measured response from the (a) TDL, (b) CRDS, and (c) FTIR, as compared with the source calibration value. The measured responses are averaged from all six tests at each source calibration value that encompassed the range of moistures and temperatures used in this evaluation.

Precision was calculated in terms of the average % relative standard deviation (RSD) at each concentration of the Unisearch S Series LasIR SM410 readings over the duration of each of the six sets included in the overall test matrix. This included six zero samples during each individual test. For each 15-min period, all 1-min readings from the SM410 were recorded, and the mean and standard deviation of those readings were calculated.

The average precision <P> was then determined as

(3)

where S is the standard deviation of the Unisearch S Series LasIR SM410 readings and X is the mean of the SM410 readings for each individual period each period, and <S/(X)> is an average over all six samples at each concentration. The average precision is reported for each of the ppmv-m source calibration values, and overall for all values is presented in Table 5. Average precision over the entire range for all six tests was 4.4% for the TDL with better precision seen for concentrations levels of 2.0 ppmv and above, 2.5% for the CRDS, and 3.5% for the FTIR.

Table 5. Average precision is reported for each of the ppmV-m source calibration values for each instrument.

HCl Source (ppmv-m)	Average Precision (%)
	TDL	CRDS	FTIR
0.5	12.9%	5.3%	6.0%
1.0	7.3%	6.1%	9.8%
2.0	3.5%	1.9%	3.2%
2.5	3.5%	1.1%	1.1%
5.0	1.9%	2.1%	1.6%
10.0	0.7%	0.9%	1.2%
25.0	1.3%	0.4%	1.4%

The minimum detection limit is reported as 3 times the average standard deviation of all the 1-min zero measurements made at each of the temperatures used in the evaluation when the instrument was evaluated over the entire range of tests. The three temperatures were 25 °C (one set of measurements, 15 points), 100 °C (three sets of measurements, 45 points), and 185 °C (six sets of measurements, 90 points). The numbers are reported in Table 6 in ppmv per meter for each instrument. The minimum detection limits for the TDL are also shown for a 5-m path, as detection limits improve with longer path lengths, and since this is the path length that is typically for power plant installations. The minimum detection limits were all on the order of 0.25 ppmv or less, utilizing the TDL values for the 5-m path. The CRDS showed the lowest MDLs of 0.06–0.07 ppmv.

Table 6. Minimum detection limits for the 3 instruments at each test temperature.

Temperature (°C)	Minimum Detection Limit (ppmv-m)
	TDL, 1 m	TDL, 5 m	CRDS	FTIR
25	0.42	0.08	0.07	0.21
100	0.86	0.17	0.06	0.25
185	1.05	0.21	0.07	0.22

Zero drift is reported in terms of the RSD of the stable readings from the daily sampling of the same HCl standard gas and zero gas supplied to the sample cell for each instrument. A total of six zero samples each containing 15 1-min samples were taken during the evaluation for each instrument. Zero drift over the six tests was found to be 1.48% for the TDL, 0.88% for the CRDS, and 1.28% for the FTIR.

Conclusions

For this study, the operational performance of three methods/instruments, including TDL spectroscopy, cavity ring down spectroscopy, and FTIR spectroscopy, were evaluated over a range of real-world operating environments. The testing was conducted in UCR’s spectroscopy evaluation laboratory. Evaluations were done over an HCl concentration range of 0–25 ppmv, and temperatures of 25, 100, and 185 °C. Tests of potential interference gases were also conducted at 185 °C for 12% H₂O, SO₂, and a matrix of CO, CO₂, and NO. A summary of the major findings and conclusions of this study are as follows:

• All instruments were sensitive to HCl to various degrees. The TDL needs at least a 5 m path length to approach sensitivities required for coal-fired power plant measurements of HCl. The CRDS was very sensitive but requires dilution and is an extractive system. The FTIR was adequate with the 8.3-m White Cell but also is an extractive system.

• The results showed good instrument accuracy as a function of temperature. The average differences with respect to temperature were 3.0% for the TDL for values over 2.0 ppmv and 6.9% of all concentrations, 3.3% for the CRDS, and 4.5% for the FTIR. The TDL was characterized separately for concentrations above 2.0 ppmv, since the 1 m path length used for these tests is less than the 5 m or more path lengths that are used at most power plant installations. These differences are roughly within the typical precision limits for the laboratory system.

• Interference tests for H₂O, SO₂, and CO, CO₂, and NO for a range of concentrations typical to flue gases in coal-fired power plants did not show strong interferences, as there was no observed effect in the system response, with the possible exception of H₂O for the FTIR. The FTIR tests results did show some increased percent differences of 10% and 10.5% (especially at the lower HCl levels) with H₂O present. The average percentage difference of 5.5% over all concentrations for the H₂O tests was also higher than the average percentage differences for the other interference tests with SO₂ and the CO, CO₂, and NO matrix. Since the FTIR is a broadband IR technique, it is always difficult to mitigate totally effects from H₂O; however, at the higher levels of HCl using a dilution system, the percent differences of the measurements were low (2.2–3.5%).

• Average precision over the entire range was 4.4% for the TDL, with better precision seen for concentrations levels of 2.0 ppmv and above, 2.5% for the CRDS, and 3.5% for the FTIR.

• The minimum detection limits were all on the order of 0.25 ppmv or less, utilizing the TDL values for the 5-m path. The CRDS showed the lowest MDLs of 0.06–0.07 ppmv.

• Zero drift was found to be 1.48% for the TDL, 0.88% for the CRDS, and 1.28% for the FTIR.

Funding

The authors acknowledge funding from the Electric Research Power Institute.

Additional information

Funding

The authors acknowledge funding from the Electric Research Power Institute.

Notes on contributors

Charles E. Dene

Charles E. Dene is a senior program manager at the Electric Power Research Institute, Palo Alto, CA.

John T. Pisano

John T. Pisano is a senior development engineer, Thomas D. Durbin is a research engineer, and Kurt Bumiller is a senior development engineer at the Bourns College of Engineering, Center for Environmental Research and Technology at the University of California at Riverside, CA.

Thomas D. Durbin

John T. Pisano is a senior development engineer, Thomas D. Durbin is a research engineer, and Kurt Bumiller is a senior development engineer at the Bourns College of Engineering, Center for Environmental Research and Technology at the University of California at Riverside, CA.

Kurt Bumiller

John T. Pisano is a senior development engineer, Thomas D. Durbin is a research engineer, and Kurt Bumiller is a senior development engineer at the Bourns College of Engineering, Center for Environmental Research and Technology at the University of California at Riverside, CA.

Keith Crabbe

Keith Crabbe is a vice-president of CEMTEK Environmental, Santa Ana, CA.

Lawrence J. Muzio

Lawrence J. Muzio is a vice-president at the Fossil Energy Research Corp., Laguna Hills, CA.