Field Measurements of Flame Scanners in a Gas-Fired Boiler under Controlled Operation Charges

Paper from the AFRC 2017 conference titled Field Measurements of Flame Scanners in a Gas-Fired Boiler under Controlled Operation Charges

Real-time monitoring of coal-fired utility boiler flames has been in active commercial implementation for at least two decades. One such system - Flame Doctor® - relies on numerical analysis and pattern matching of flickering dynamics as measured by visible-IR flame scanners installed on each burner. This diagnosis system can monitor up to hundreds of burners in large furnaces and allow operators to optimize the unit for combustion efficiency and emissions. Field experience has shown that significant improvements in emissions performance can be achieved with changes to just a small number of burners, and the ability to monitor all flames, especially where visual access is not possible, is of great utility for plant operation.; We have begun adapting the successful Flame Doctor paradigm to gas-fired furnaces and process heater systems. As a first step, we present here analysis of field data acquired from a six-burner gas-fired furnace firing natural gas. A portable Flame Doctor system monitored the output of custom visible-UV flame scanners, and the furnace operating parameters were varied through a series of controlled experiments. We show how the statistical changes in the furnace flame response show promise for a gas-fired flame-monitoring system which would be applicable from single-burner process heaters to large, multi-burner furnaces.

Finney/Fuller/Flynn/Daw - Field measurements of flame scanners in a gas-fired boiler under controlled operating changes AFRC 2017 Industrial Combustion Symposium - Houston Field measurements of flame scanners in a gas-fired boiler under controlled operating changes † Charles E.A. Finney1,*, Timothy A. Fuller2, Thomas J. Flynn2, C. Stuart Daw1 1 2 Oak Ridge National Laboratory, Oak Ridge TN 37831 USA The Babcock & Wilcox Company, Barberton OH 44203 USA Real-time monitoring of coal-fired utility boiler flames has been in active commercial implementation for at least two decades. One such system - Flame Doctor® - relies on numerical analysis and pattern matching of flickering dynamics as measured by visible-IR flame scanners installed on each burner. This diagnosis system can monitor up to hundreds of burners in large furnaces and allow operators to optimize the unit for combustion efficiency and emissions. Field experience has shown that significant improvements in emissions performance can be achieved with changes to just a small number of burners, and the ability to monitor all flames, especially where visual access is not possible, is of great utility for plant operation. We have begun adapting the successful Flame Doctor paradigm to gas-fired furnaces and process heater systems. As a first step, we present here analysis of field data acquired from a six-burner gas-fired furnace firing natural gas. A portable Flame Doctor system monitored the output of custom visible-UV flame scanners, and the furnace operating parameters were varied through a series of controlled experiments. We show how the statistical changes in the furnace flame response show promise for a gas-fired flame-monitoring system which would be applicable from single-burner process heaters to large, multi-burner furnaces. Motivation and overall objective‡ Utilities with coal-fired boilers are under increasing scrutiny and pressure to improve efficiency, thereby lowering greenhouse gas emissions, and to reduce pollutant emissions. The past two decades has seen significant progress in real-time monitoring of boiler flame and emissions performance, with several technologies currently in commercial implementation. In field testing during development of the Flame Doctor® system, it was demonstrated how just one flame out of dozens in a boiler could severely affect overall boiler emissions and that having simple diagnostics of flame stability and state could enable boiler operators to make targeted changes with significant improvements, even when they lacked direct visual access to the poor-performing flames [Fuller (2004)]. While gas-fired combustion, considered cleaner than coal-fired combustion, has received less regulatory attention, it is now an opportune time to prepare for future constraints and the next generation of gas-fired systems. Domestically, with the conversion of many coal-fired boilers to natural gas, there is an opportunity to provide a means to "tune" a retrofitted natural gas Notice: This manuscript has been authored by UT-Battelle, LLC, under Contract No. DE-AC0500OR22725 with the U.S. Department of Energy. The United States Government retains and the publisher, by accepting the article for publication, acknowledges that the United States Government retains a non-exclusive, paid-up, irrevocable, world-wide license to publish or reproduce the published form of this manuscript, or allow others to do so, for the United States Government purposes. The Department of Energy will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan (http://energy.gov/downloads/doe-public-access-plan). * Corresponding author: FINNEYC @ORNL.GOV or 865-946-1243. † ‡ Note: Significant components of the introductory and methodology descriptive text herein are included verbatim from our paper from last year's AFRC meeting [Finney (2015, 2016)]. We choose this approach of verbatim inclusion over reference to make this paper self-contained in its narrative, in case of limited availability of the cited paper. Compared with last year's paper, the present paper changes focus on application, analytical methodology, results and conclusions. Page 1 of 10 Finney/Fuller/Flynn/Daw - Field measurements of flame scanners in a gas-fired boiler under controlled operating changes AFRC 2017 Industrial Combustion Symposium - Houston combustion system to a boiler cavity originally optimized around pulverized coal combustion, and to maintain it at near-optimal performance. Globally, countries that have not had strict emissions control regulations are now implementing strict emissions regulations as their utility and industrial capacity in fossil fuel based units grows, and international pressure on greenhouse gas emissions strengthens. The state of the art in boiler flame monitoring has moved from time-averaged, usually linear, metrics to nonlinear metrics which account for the dynamical behavior of flames. Capturing both long- and short-timescale dynamical features of flames is important because many unstable and transient behaviors contribute disproportionately to poor combustion and emissions performance, and time-averaged metrics often do not quantify such events. With the continued development of hardware and software systems amenable for optical flame sensing in boiler furnaces, we are planning for the next generation of optical flame diagnostic systems. Continuous monitoring of burner operation has become increasingly important toward achieving and maintaining optimum performance as incremental hardware improvements in combustion system design have become more difficult to achieve. In this paper, we present results from a field test to explore the feasibility of applying an approach such as implemented in Flame Doctor (developed for coal-fired burners) to a gas-fired boiler. An ultimate objective is to implement whole-flame analysis for gas-fired burners, but in the interim, an approach using spatially localized flame-scanner measurements might suffice and could tie in with existing flame-scanner monitoring systems currently used in coal-fired boilers, as well as with any safety or controls systems in the gas-fired unit. Brief overview of prior work The following discussion focuses on optical measurements of flames for diagnostic purposes. Other systems using non-optical measurements, such as with pressure, acoustic, or ionic sensors, or which do not directly measure flame properties, such as with flue-gas spatial analyzers, are not included. Most commercially offered flame-monitoring systems have focused on coal-fired flames, but their analytical methodology generally could be extended to oil- and gas-fired systems with appropriate measurements. While several ad-hoc schemes for measuring whole-flame images or signals from light-intensity sensors had been utilized informally for decades, in the mid-1990s commercial systems began to be developed which focused on statistical analysis of flame measurements. One such system was an optical system which used the Fourier (frequency-based) spectrum of flame flicker to assign a numerical quality metric [Khesin (1996a, 1996b, 1997)]. The Fourier spectrum is linear and captures first-order effects on a time-averaged basis, but it does not capture all the nuances of dynamical variations seen in nonlinear systems, such as flames in certain unstable combustion regimes such as with staged combustion of pulverized coal burners. Concurrently, techniques derived from the study of nonlinear dynamical and chaotic systems began to be applied to boiler flames [Fuller (1996a, 1996b)]. Continued development under collaboration of the Electric Power Research Institute, The Babcock & Wilcox Company and Oak Ridge National Laboratory led to the development of the Flame Doctor system for commercial release in 2004 [Flynn (2003); Daw (2003)]. The Flame Doctor system utilizes flame-scanner signals, typically focused on a relatively small volume of the flame seen through a 10-15 cm (4-6 inch) diameter and several meter long sight tube, near the burner nozzle, and applies a series of statistical analyses of linear and nonlinear metrics to determine the relative stability of the flame. By using existing flame scanners installed for safety purposes or using custom scanners, the system can be quickly implemented on most utility boilers. The Flame Doctor system has been used on a temporary or permanent basis on dozens of utility boilers in both the US and internationally and has seen a good Page 2 of 10 Finney/Fuller/Flynn/Daw - Field measurements of flame scanners in a gas-fired boiler under controlled operating changes AFRC 2017 Industrial Combustion Symposium - Houston degree of success in identifying poorly performing burners. Correction of the boiler operation, largely in supervised open-loop control, either on a burner or mill-group basis has yielded very favorable improvements in boiler emissions (such as NOx and CO) by improving combustion quality. While this experience has had considerable value in improving boiler performance, ever more restrictive Environmental Protection Agency (EPA) boiler emissions limits demonstrate that enhanced optical sensors and improved dynamical characterization of flame, mill-group and overall boiler dynamics, reflecting the latest understanding of nonlinear dynamical systems theory, will be worth the development effort to achieve even greater improvement in boiler performance. Additionally, the latest EPA regulations recognize the importance of advanced closed-loop control systems; therefore, the integration of advanced sensors into these control systems will continue to be a priority. Results from a field feasibility test We focus on a single plant test from an operating gas-fired boiler system. The objective was to see whether flame-scanner signals statistically tracked controlled operating changes. This test was designed to provide experimental verification to support future research into such flame-scanner, and later imaging, monitoring and control systems in gas-fired furnaces and boilers. Field test site The test unit was an operating utility boiler firing natural gas, with a full-load capacity of ~24 MWe, and data were collected by Babcock & Wilcox personnel during a site visit. There were six wallflush burners arranged in two stacked rows of three, and the outside two columns of burners were instrumented for flame-signal recording. A portable Flame Doctor unit, with standard signal conditioning of amplification and bandpass analog filtering, was connected to four temporarily installed, custom visible-UV sensitive flame scanners which were mounted to sight through wall access ports behind the four selected burners. Visual access to the flames and firebox was available through a wall port for making observations of flame extent, shape and quality. Additionally, unit operating variables such as air and fuel flow rates, burner settings, and performance, as well as NOx and CO emissions measurements, were recorded. A series of tests was performed in sequence over several days. Scanner analog voltage signals were digitized continuously at 1000 Hz (highpass filtered at 0.1 Hz to remove equipment drift and lowpass filtered at around 450 Hz to minimize signal aliasing) in 120 sec increments during a series of controlled changes to the boiler/burner system. Typically, each test condition involved ~30 minutes of steady-state operation following 15 minutes of settling time after controlled changes; data here are analyzed only from the steady-state portions of each test condition. Because the flame-scanner sighting was temporary and ad hoc using available access ports, the viewing angle was not necessarily optimal but generally observed the flame. In some instances, the flame moved out of the scanner sighting for intermittent or extended periods, resulting in absent or degraded signals; these instances were excluded from analysis because of the skewing effect on statistics and metrics of interest (for instance, the probability distribution of the signals would be highly non-Gaussian or outside the range of typically flame-scanner signals). One scanner, designated #1, is omitted from analysis herein because of this problem, and the other three scanners only had one time series each omitted because of sighting issues. Page 3 of 10 Finney/Fuller/Flynn/Daw - Field measurements of flame scanners in a gas-fired boiler under controlled operating changes AFRC 2017 Industrial Combustion Symposium - Houston Analysis techniques As a first-pass proof of concept, we here analyze up to three recorded time series (unless excluded, as described above) at four test cases for three scanners. Here, we note response to changes in the scanner signals to changes in test conditions without attempting to ascribe causes such as mixing, stoichiometry etc. The following analytical techniques were employed: • Signal oscillograms - Depiction of the qualitative patterns in the raw signals • Higher-order moments (skewness & kurtosis) - Measures of data probability distributions • Temporal irreversibility (T3) functions - Measure of time asymmetry in signal oscillations Skewness is the third statistical moment (after mean and variance) and is defined as: ⟨(𝑥𝑖 − 𝑥)3 ⟩ 𝜎3 where ⟨∙⟩ denotes the expected value, xi is the value of the time series at index i, 𝑥 = ⟨𝑥𝑖 ⟩ is the data mean, and 𝜎 is the data standard deviation. 𝑆= Kurtosis is the fourth statistical moment and is defined as: ⟨(𝑥𝑖 − 𝑥)4 ⟩ 𝐾= −3 𝜎4 where the subtraction of 3 shifts K to 0 for a Gaussian distribution. The temporal irreversibility function is defined as: 𝑇3 (𝑘) = √𝑁 3 ∑𝑖(∆𝑖,𝑘 ) 3 2 2 (∑𝑖(∆𝑖,𝑘 ) ) where ∆𝑖,𝑘 = 𝑥𝑖+𝑘 − 𝑥𝑖 and k is a temporal displacement (lag) from i and N is the number of differences applied in ∆𝑖,𝑘 . The denominator normalizes T3, which can have either negative or positive values, and calculation over a range of lags k helps quantify temporal relationships such as is common with correlation functions. These metrics have been used successfully in nonlinear time series analysis of measured flame dynamics and are integral parts of the Flame Doctor system. Page 4 of 10 Finney/Fuller/Flynn/Daw - Field measurements of flame scanners in a gas-fired boiler under controlled operating changes AFRC 2017 Industrial Combustion Symposium - Houston Example signals and their statistical changes Figure 1. Example time series for three scanners at the four test conditions. The timespan of each plot is 120 s, and the ordinate is scaled to the data range. Representative time series at each test condition for three of the flame scanners of interest are plotted in Figure 1. At the scales plotted, small-scale differences are not fully obvious, but these will be characterized quantitatively. Large-scale differences include the degree of spikeyness and small bursts of either abnormally large or small variance. Page 5 of 10 Finney/Fuller/Flynn/Daw - Field measurements of flame scanners in a gas-fired boiler under controlled operating changes AFRC 2017 Industrial Combustion Symposium - Houston Figure 2. Data skewness for the three burners, with multiple measurements of skewness (ordinate) per test condition (abscissa). All axes are the same scale. The first metric of interest is the data skewness, which is a measure of how the data probability density function is situated relative to the mean. Previous analysis of flame-scanner data has shown that skewness sometimes relates to flame-flicker dropout, such as might occur with the flame lifting out of the field of view and then slowly reattaching. Skewness values for the three burners and four tests examined are shown in Figure 2. At each test condition, there are multiple skewness values calculated from the replicate time series recorded at each point; these replicates help define the repeatability or scatter in the data at nominally the same condition. The overall picture is mixed - for some burners and some tests, there are discernible statistical differences, whereas sometimes there are not. The complete story is told in the ensemble of statistics, but in this single measure, the scanner signals do change with test condition. Page 6 of 10 Finney/Fuller/Flynn/Daw - Field measurements of flame scanners in a gas-fired boiler under controlled operating changes AFRC 2017 Industrial Combustion Symposium - Houston Figure 3. Data kurtosis for the three burners, with multiple measurements of kurtosis (ordinate) per test condition (abscissa). All axes are the same scale. The second metric of interest is the data kurtosis, which is a measure of how the data probability density function is shaped, especially relative to a Gaussian distribution. Previous analysis of flamescanner data has shown that kurtosis sometimes relates to the short-term signal spikes reflective of flame edge-lifting from the burner throat, among other causes. Kurtosis values for the three burners and four tests examined are shown in Figure 3. At each test condition, there are multiple kurtosis values calculated from the replicate time series recorded at each point; these replicates help define the repeatability or scatter in the data at nominally the same condition. As with skewness, the overall picture is mixed - for some burners and some tests, there are discernible statistical differences, whereas sometimes there are not. The complete story is told in the ensemble of statistics, but in this single measure, the scanner signals do change with test condition. Page 7 of 10 Finney/Fuller/Flynn/Daw - Field measurements of flame scanners in a gas-fired boiler under controlled operating changes AFRC 2017 Industrial Combustion Symposium - Houston Figure 4. Example T3 functions for three burners at the four test conditions. The timespan of lags in each plot is 1 s, and the ordinate is on the same scale for all functions; the 0-axis is highlighted. Another metric of interest is the T3 function, which characterizes whether a statistical description of a signal varies depending on whether it is analyzed in a forward- or reverse-time sense. Significant temporal irreversibility is important because Gaussian sources, or static transforms thereof, are temporally symmetric, and irreversibility is a hallmark of (but not sufficient condition for) nonlinear signals. The T3 function measures the degree of irreversibility over a range of timescales, much like is done with correlation functions. Previous analysis of flame-scanner data has shown that T3 relates to combustion or flame instabilities, as opposed to flow instabilities. T3 functions for the three burners and four tests examined are shown in Figure 4. There are some consistencies amond channels and tests, and the most prominent difference is seen in the final test, where T3 drops close to 0 over all timescales. The trends in the shapes of the functions as a function of test condition give confidence that the flame scanner signals are responding well to changes in operating conditions. Page 8 of 10 Finney/Fuller/Flynn/Daw - Field measurements of flame scanners in a gas-fired boiler under controlled operating changes AFRC 2017 Industrial Combustion Symposium - Houston Conclusions and recommendations Currently implemented boiler diagnostic systems for coal flames have shown that real-time monitoring of combustion performance, on a per-flame and/or mill-group basis, can yield significant improvements in combustion quality and thus improve emissions and increase efficiency. Extension to natural-gas flames in some respects offers greater opportunities because whole-flame optical measurements are more easily made than within coal-fired furnaces. As an interim step, the feasibility of using a flame-scanner approach is attactive, especially to tie into any existing plant hardware and to account for cases in which flame sighting is not available. This work presented in this paper has determined that using flame scanners, sensitive to the visible-UV spectrum, is a feasible approach and that adapting existing systems, such as Flame Doctor, is justified. In followup work, we will use the entire available data record and attempt to correlate changes in metrics with specific control changes in the test conditions. This will be a necessary step in producing a final, viable system which can operate flexibly in a diverse environment of furnaces and boilers. Acknowledgements Babcock & Wilcox and Oak Ridge National Laboratory acknowledge the support of the Electric Power Research Institute during the development of the Flame Doctor system. We also recognize and acknowledge the contributions of engineers from Alliant Energy, Ameren, and Southern Company during the initial field trials of the Flame Doctor system. Portions of this work (signal analysis) were supported by the U.S. Department of Energy's (DOE) Office of Energy Efficiency and Renewable Energy, Advanced Manufacturing Office's (AMO), Combined Heat and Power Research Program, and performed at Oak Ridge National Laboratory (ORNL) by UT-Battelle, LLC, under Contract No. DE-AC05-00OR22725. The authors thank Bob Gemmer of the AMO for his support. References DAW S, FINNEY C, BAILEY R, FLYNN T, FULLER T (2002). Real-time monitoring of dynamical state changes in staged combustion. In ASME International Mechanical Engineering Congress and Exposition, Proceedings of the ASME Heat Transfer Division, Paper 39053. DAW CS, FINNEY CEA, FULLER TA, FLYNN TJ, BAILEY RT (2003). Real-time monitoring of flame quality in staged coal combustion with the Flame Doctor™ system. 2003 American Flame Research Committee International Symposium on Combustion Research and Industrial Practice: Bridging the Gap (Livermore, California USA; 2003 October 16-17). FINNEY CEA, DAW CS, FULLER TA, FLYNN TJ, KULP CW (2015). Opportunities for the next generation of optical boiler diagnostics. 2015 American Flame Research Committee Industrial Combustion Symposium (Salt Lake City, Utah USA; 2015 September). FINNEY CEA, KULP CW, DAW CS, ALAVANDI S, FULLER TA, FLYNN TJ, KULP CW (2016). Opportunities for optical flame diagnostics in commercial and industrial furnaces. 2016 American Flame Research Committee Industrial Combustion Symposium (Koloa, Hawai'i USA; 2016 September). FLYNN T, BAILEY R, FULLER T, DAW CS, FINNEY CEA, STALLINGS J (2003). Flame monitoring enhances burner management. Power Engineering, February 2003: 50-54. FULLER TA, FLYNN TJ, DAW CS (1996a). Analysis of dynamic boiler measurements: a practical approach. The Chemical Engineering Journal 64: 179-189. Page 9 of 10 Finney/Fuller/Flynn/Daw - Field measurements of flame scanners in a gas-fired boiler under controlled operating changes AFRC 2017 Industrial Combustion Symposium - Houston FULLER TA, FLYNN TJ, DAW CS (1996b). Enhancing burner diagnostics and control with chaos-based signal analysis techniques. In ASME International Mechanical Engineering Congress and Exposition, Proceedings of the ASME Heat Transfer Division 4: 281-291. FULLER TA, FLYNN TJ, BAILEY RT, DAW CS, FINNEY CEA, HUTCHINSON D, MUSGROVE B, LASSAHN R, STALLINGS J (2004). Advances in utility applications of the Flame Doctor system. EPRI-DOE-EPA-AWMA Combined Power Plant Air Pollutant Control Mega Symposium (2004 August 30 - September 2; Washington DC). Available at http://www.babcock.com/library/Documents/br-1753.pdf. FULLER TA, DAW CS, FINNEY CEA (2014). Burner performance monitoring using advanced analysis of optical signals. Power-Gen Europe (2014 June 3-5; Cologne, Germany). Available at http://www.babcock.com/library/Documents/BR-1908.pdf. KHESIN M, SENIOR C, ELDREDGE T, LEVY E, SARUNAC N, BELLES D, MARTH R (1996a). Application of a flame spectra analyzer for burner balancing. Proceedings of the 6th International ISA POWID/EPRI Controls & Instrumentation Conference (Baltimore, Maryland; June 1996). KHESIN MJ (1996b). Demonstration of new frequency-based flame monitoring system. Proceedings of the 58th Annual Meeting of the American Power Conference 58-II: 1010. KHESIN M, GIRVAN R, QUENAN D (1997). Demonstration tests of a new burner diagnostic system on a 650 MW coal-fired utility boiler. Proceedings of the 59th Annual Meeting of the American Power Conference 59-I: 325-330. Page 10 of 10