222 lines
11 KiB
TeX
222 lines
11 KiB
TeX
\documentclass[11pt,a4paper]{article}
|
|
|
|
\usepackage[margin=25mm]{geometry}
|
|
\usepackage{amsmath}
|
|
\usepackage{amssymb}
|
|
\usepackage{booktabs}
|
|
\usepackage{hyperref}
|
|
\usepackage{siunitx}
|
|
\usepackage{enumitem}
|
|
|
|
\hypersetup{
|
|
colorlinks=true,
|
|
linkcolor=blue,
|
|
urlcolor=blue,
|
|
citecolor=blue
|
|
}
|
|
|
|
\title{Strengths and Weaknesses of Flame Chemiluminescence Diagnostics\\
|
|
\large with Notes on Ammonia-Flame Radical Signatures}
|
|
\author{Robin Clark}
|
|
\date{\today}
|
|
|
|
\begin{document}
|
|
\maketitle
|
|
|
|
\section{Central Argument}
|
|
|
|
The most useful way to understand OH* and CH* chemiluminescence is this:
|
|
|
|
\begin{quote}
|
|
Chemiluminescence measures the active flame reaction zone, not the fully mixed and cooled exhaust products in the flue.
|
|
\end{quote}
|
|
|
|
That one statement explains both the strength and the weakness of the method. It is strong because it gives a fast, direct view of combustion while it is happening. It is weak because final flue-gas composition depends on the whole downstream history of the gases, including mixing, residence time, quenching, dilution and post-flame oxidation.
|
|
|
|
Therefore, OH*/CH* chemiluminescence should not normally be treated as a direct replacement for a zirconia oxygen probe or a CO analyser. Its better role is as a dynamic combustion-process diagnostic.
|
|
|
|
\section{What OH* and CH* Actually Tell Us}
|
|
|
|
Common flame chemiluminescence bands include approximately:
|
|
|
|
\begin{center}
|
|
\begin{tabular}{lll}
|
|
\toprule
|
|
Species & Approximate band & Typical interpretation \\
|
|
\midrule
|
|
OH* & \SI{306}{\nano\metre}--\SI{310}{\nano\metre} & oxidation / high-temperature reaction zone \\
|
|
CH* & about \SI{430}{\nano\metre} & hydrocarbon reaction-zone activity \\
|
|
C$_2$* Swan bands & about \SI{516}{\nano\metre} and nearby bands & carbon-rich hydrocarbon chemistry \\
|
|
\bottomrule
|
|
\end{tabular}
|
|
\end{center}
|
|
|
|
The ratio OH*/CH* is often used as an indicator of equivalence-ratio trend in controlled flames. However, in industrial burners the calibration is strongly dependent on burner geometry, fuel, air staging, optical path and flame structure.
|
|
|
|
\section{Strengths of Chemiluminescence}
|
|
|
|
\subsection{Very Fast Response}
|
|
|
|
The light emission comes from short-lived excited radicals. The optical signal therefore responds on very short timescales, often far faster than immersed or extractive gas sensors. This makes the method well suited to detecting combustion dynamics, transients and oscillatory behaviour.
|
|
|
|
By contrast, flue-gas sensors involve gas transport, diffusion, sample extraction, thermal equilibration or electrochemical response. These are naturally slower.
|
|
|
|
\subsection{Direct Observation of the Flame Front}
|
|
|
|
A flue oxygen probe measures the result of combustion after the gases have mixed and moved downstream. Chemiluminescence observes the active reaction zone itself. This means it can detect local flame phenomena that may be invisible in a common flue measurement, such as:
|
|
|
|
\begin{itemize}[itemsep=2pt]
|
|
\item flame lift-off or attachment changes;
|
|
\item local rich or lean pockets;
|
|
\item poor premixing;
|
|
\item recirculation-zone changes;
|
|
\item burner-to-burner imbalance;
|
|
\item combustion intensity changes.
|
|
\end{itemize}
|
|
|
|
\subsection{Excellent for Instability and Surge Detection}
|
|
|
|
The time-varying part of the OH* or CH* signal can be more valuable than its mean value. Useful quantities include:
|
|
|
|
\begin{itemize}[itemsep=2pt]
|
|
\item standard deviation of OH* intensity;
|
|
\item RMS fluctuation level;
|
|
\item FFT peak frequencies;
|
|
\item phase relationship between optical channels;
|
|
\item amplitude modulation during load changes.
|
|
\end{itemize}
|
|
|
|
These quantities can reveal thermoacoustic oscillation, unstable premixing, pulsation, flame flicker changes, blow-off approach or surging. This is probably one of the strongest technical uses of the method.
|
|
|
|
\subsection{Non-Contact Sensing}
|
|
|
|
The detector can view the flame through a window, sight tube, quartz optic or sapphire optic. This avoids placing a fragile probe directly into hot, corrosive or dirty flue gas. In industrial settings this is a real advantage, provided the optical path can be kept clean or monitored for fouling.
|
|
|
|
\subsection{Possible Fuel Fingerprinting}
|
|
|
|
Changing the fuel changes the radical population and reaction pathways. For example, hydrogen addition, ammonia addition or syngas variation can alter the balance of OH*, CH*, NH*, NH$_2$*, C$_2$* and related emissions.
|
|
|
|
This suggests a useful role as a combustion fingerprint sensor, especially when combined with conventional O$_2$, NOx or CO measurements.
|
|
|
|
\section{Weaknesses of Chemiluminescence}
|
|
|
|
\subsection{It Does Not Directly Measure Final Products}
|
|
|
|
This is the main limitation. OH* and CH* exist in the active flame region, but final stack O$_2$ and CO depend on events after the flame front:
|
|
|
|
\begin{itemize}[itemsep=2pt]
|
|
\item post-flame oxidation of CO to CO$_2$;
|
|
\item residence time at high temperature;
|
|
\item mixing with secondary air;
|
|
\item wall quenching;
|
|
\item dilution and leakage air;
|
|
\item staged combustion;
|
|
\item recirculation and burnout length.
|
|
\end{itemize}
|
|
|
|
Two flames with similar OH*/CH* ratios can therefore produce different downstream O$_2$ or CO concentrations.
|
|
|
|
\subsection{Strong Geometry Dependence}
|
|
|
|
The measured signal depends on where the sensor looks and what part of the flame is in view. Burner shape, viewing angle, optical path length, flame luminosity and sight-tube fouling can all change the measured signal without a corresponding change in the true global combustion state.
|
|
|
|
\subsection{Local Rather Than Global Measurement}
|
|
|
|
This is both an advantage and a weakness. A local optical measurement can detect a local fault quickly, but it may not represent the overall boiler exhaust. For multi-burner systems this can be an advantage if each burner is monitored separately, but it is a weakness if one tries to infer whole-stack gas composition from one optical view.
|
|
|
|
\subsection{Optical Fouling and Sensor Drift}
|
|
|
|
Industrial combustion environments can produce soot, ash, refractory dust, condensate and window contamination. These can attenuate UV and visible light differently. Detector aging and window fouling can therefore mimic chemistry changes unless the instrument includes compensation or self-checking.
|
|
|
|
\subsection{Fuel Dependence}
|
|
|
|
A calibration developed for methane may not transfer to hydrogen, propane, oil, ammonia blends or syngas. This is particularly important for alternative fuels where radical chemistry changes substantially.
|
|
|
|
\section{Best Engineering Position}
|
|
|
|
The strongest claim is not:
|
|
|
|
\begin{quote}
|
|
``Chemiluminescence accurately measures residual oxygen and carbon monoxide.''
|
|
\end{quote}
|
|
|
|
The stronger and more defensible claim is:
|
|
|
|
\begin{quote}
|
|
``Chemiluminescence gives a fast, non-contact, reaction-zone diagnostic of flame state, stability, dynamics, burner balance and fuel-dependent combustion behaviour.''
|
|
\end{quote}
|
|
|
|
This allows the technology to complement conventional instruments rather than pretending to replace them.
|
|
|
|
\section{A Practical Sensor-Fusion View}
|
|
|
|
A sensible industrial architecture would be:
|
|
|
|
\begin{center}
|
|
\begin{tabular}{lll}
|
|
\toprule
|
|
Measurement & Best at & Weak at \\
|
|
\midrule
|
|
OH*/CH* optics & flame-front state, dynamics, instability & exact stack O$_2$/CO \\
|
|
Zirconia O$_2$ probe & residual oxygen after burnout & fast local flame dynamics \\
|
|
CO analyser & incomplete combustion product measurement & fast flame-front diagnosis \\
|
|
NOx analyser & integrated NOx production outcome & local flame stability \\
|
|
IR/UV flame scanner & flame presence / safeguard function & rich chemical diagnostics \\
|
|
\bottomrule
|
|
\end{tabular}
|
|
\end{center}
|
|
|
|
The strongest product concept is therefore an augmented flame-quality instrument: optical flame diagnostics plus conventional gas sensing where absolute exhaust-gas concentrations are required.
|
|
|
|
\section{Ammonia Combustion: UV and Visible Radical Signatures}
|
|
|
|
Ammonia combustion introduces nitrogen-containing radical chemistry. Published ammonia-flame chemiluminescence work commonly reports UV and visible signatures from species such as NO*, OH*, NH* and NH$_2$*.
|
|
|
|
Important reported bands include approximately:
|
|
|
|
\begin{center}
|
|
\begin{tabular}{lll}
|
|
\toprule
|
|
Species & Approximate region & Comment \\
|
|
\midrule
|
|
OH* & about \SI{309}{\nano\metre} & still an important high-temperature oxidation marker \\
|
|
NH* & about \SI{336}{\nano\metre} & nitrogen-hydrogen radical signature in UV \\
|
|
NH$_2$* & visible, often red/orange bands around \SI{600}{\nano\metre}--\SI{650}{\nano\metre} & contributes to visible ammonia-flame colour \\
|
|
NO* & UV region & related to nitrogen oxide formation chemistry \\
|
|
NO$_2$* & visible broadband contribution, especially in lean post-flame regions & can affect blue/visible colour balance \\
|
|
CH* & about \SI{430}{\nano\metre} & present when hydrocarbon co-fuels such as methane are present \\
|
|
CN* & UV/visible bands in some ammonia/hydrocarbon cases & possible carbon--nitrogen chemistry marker \\
|
|
\bottomrule
|
|
\end{tabular}
|
|
\end{center}
|
|
|
|
The most useful point is that ammonia flames may provide additional optical channels beyond OH* and CH*. In particular, NH* and NH$_2$* can be useful markers of ammonia-related reaction chemistry, while NO*/NO$_2$* emissions can relate to nitrogen oxide pathways.
|
|
|
|
However, these optical signatures should still be treated as reaction-zone and near-flame diagnostics. They do not directly provide certified downstream NOx, O$_2$, CO, N$_2$O or unburnt NH$_3$ slip concentrations without calibration and validation.
|
|
|
|
\section{Implication for an Industrial Instrument}
|
|
|
|
For hydrocarbon flames, OH*/CH* may be useful for rich/lean trend and stability. For ammonia, ammonia/hydrogen, ammonia/methane or syngas blends, the optical approach could be extended to include:
|
|
|
|
\begin{itemize}[itemsep=2pt]
|
|
\item OH* near \SI{309}{\nano\metre};
|
|
\item NH* near \SI{336}{\nano\metre};
|
|
\item CH* near \SI{430}{\nano\metre}, where hydrocarbon fuel is present;
|
|
\item NH$_2$* visible bands around the orange/red region;
|
|
\item C$_2$* Swan bands for carbon-rich or hydrocarbon/syngas flames;
|
|
\item time-domain statistics and FFT analysis for instability.
|
|
\end{itemize}
|
|
|
|
This supports the idea of a multi-channel combustion-state sensor. The most credible use would be fast flame-quality and instability monitoring, with possible fuel-composition fingerprinting. Absolute gas concentration measurement should remain the role of validated gas sensors.
|
|
|
|
\section{References and Starting Points}
|
|
|
|
\begin{itemize}
|
|
\item B. Cosway et al., work on ammonia/hydrogen flame structure and NO production; reports NO*, OH*, NH* in the UV and NH$_2$* in the visible for ammonia flames.
|
|
\item S. Mashruk et al., studies of ammonia/hydrogen and ammonia/methane flame chemiluminescence and emissions.
|
|
\item X. Zhu et al., UV-visible chemiluminescence signatures of laminar ammonia-hydrogen-air flames.
|
|
\item W. Weng et al., visible chemiluminescence of ammonia premixed flames, including NH$_2$* and NO$_2$* visible-region effects.
|
|
\end{itemize}
|
|
|
|
\end{document}
|