**4. Harsh environment applications and examples**

The applications of TDLAS and QCLAS techniques in harsh environment conditions are well known and widely discussed in the literature. The most common applications include measurement of trace gas concentrations [44–46], temperature [19, 47], combustion control [48], and plasma diagnostics [49]. It is beyond the scope of this chapter to discuss all these applications in detail. In the following subsections, two specific examples will be discussed which demonstrate the power of this technology in tackling real‐world challenges where conventional techniques fail to perform well.

#### **4.1. Steam quality sensor for steam turbine applications**

This subsection is a summarized excerpt of the work published in reference [22]. Steam quality or steam wetness fraction is a critical operational parameter in steam turbines and is used in estimation of turbine efficiency and remaining life. It is a quantitative measure of the amount of water vapor and liquid water (usually microdroplets) present in the process steam. Steam quality or wetness fraction (*X*) is defined as follows

$$X = \frac{m\_{\text{vapor}}}{m\_{\text{vapor}} + m\_{\text{liquid}}} \tag{5}$$

where mvapor and mliquid are the mass of vapor phase and mass of liquid phase, respectively.

The steam quality in a steam turbine needs to be closely monitored as an indicator of wear and tear of the machinery. The architecture of a steam turbine mainly includes three sections: HP (high pressure), IP (intermediate pressure), and LP (low pressure) respectively. The steam in HP and IP sections has little liquid water content, and therefore, the efficiency of these sections can be measured using standard temperature and pressure measurements. The LP section, however, can have steam with significant liquid water content (also known as wet steam), which can cause significant erosional damage to the rotating components [50, 51]. Therefore, the accurate measurement of steam wetness in the LP section becomes extremely critical. Most conventional methods are either based on steam sampling or based on calculations involv‐ ing machine parameters (such as power output), both of which have low accuracy [22]. Therefore, a diode laser‐based sensor capable of direct in situ measurement of steam quality in the hot and harsh LP section would be an ideal solution to this challenge.

applications involving multiple lasers, the use of hardware lock‐in amplifiers can be cumber‐ some and bulky. In such cases, an onboard PC/processor with software lock‐in feature can be a much better solution. It should be noted that software lock‐in features are commonly implemented in development environments such as National Instruments Labview and MATLAB. This concludes the basic overview of TDLAS sensor design and the following section will discuss some examples of how this technology is enabling real‐world solutions to

The applications of TDLAS and QCLAS techniques in harsh environment conditions are well known and widely discussed in the literature. The most common applications include measurement of trace gas concentrations [44–46], temperature [19, 47], combustion control [48], and plasma diagnostics [49]. It is beyond the scope of this chapter to discuss all these applications in detail. In the following subsections, two specific examples will be discussed which demonstrate the power of this technology in tackling real‐world challenges where

This subsection is a summarized excerpt of the work published in reference [22]. Steam quality or steam wetness fraction is a critical operational parameter in steam turbines and is used in estimation of turbine efficiency and remaining life. It is a quantitative measure of the amount of water vapor and liquid water (usually microdroplets) present in the process steam. Steam

> *vapor vapor liquid*

where mvapor and mliquid are the mass of vapor phase and mass of liquid phase, respectively.

The steam quality in a steam turbine needs to be closely monitored as an indicator of wear and tear of the machinery. The architecture of a steam turbine mainly includes three sections: HP (high pressure), IP (intermediate pressure), and LP (low pressure) respectively. The steam in HP and IP sections has little liquid water content, and therefore, the efficiency of these sections can be measured using standard temperature and pressure measurements. The LP section, however, can have steam with significant liquid water content (also known as wet steam), which can cause significant erosional damage to the rotating components [50, 51]. Therefore, the accurate measurement of steam wetness in the LP section becomes extremely critical. Most conventional methods are either based on steam sampling or based on calculations involv‐ ing machine parameters (such as power output), both of which have low accuracy [22].

*<sup>m</sup> <sup>X</sup> m m* <sup>=</sup> <sup>+</sup> (5)

challenging industrial problems.

404 406High Energy and Short Pulse Lasers

conventional techniques fail to perform well.

**4.1. Steam quality sensor for steam turbine applications**

quality or wetness fraction (*X*) is defined as follows

**4. Harsh environment applications and examples**

As shown in Eq. (5), estimation of steam quality requires a quantitative measurement of water in liquid and vapor phases. **Figure 4** shows the schematic of a steam quality measurement system designed using two fixed wavelength broadband diode lasers. A laser at 945 nm (power 20 mW, spectral width ∼2 nm) is used to probe the absorption of water vapor. Similarly, a laser at 1560 nm (20 mW, spectral width ∼10 nm) is used for liquid water. Both lasers are passed through a steam chamber, which is capable of generating known vapor/liquid ratios. The path length of the steam chamber is predetermined. The intensities of both lasers before and after passage through the steam chamber are measured with Indium Gallium Arsenide (InGaAs) detectors. A comparison of the intensity loss is used to generate a quanti‐ tative measurement of the vapor and liquid concentrations respectively and hence, to estimate the steam quality, according to Eqs. (1) and (4).

**Figure 4.** Schematic of the experimental setup for steam quality measurement. Figure taken with permission from IEEE (From our published paper).

**Figure 5a** and **b** shows a comparison of the laser‐based steam quality measurement with the mass flow rate‐based method under different water spray conditions. A reasonably good match is found between the two techniques and further improvements are possible through improvement of analysis algorithms for different temperature, pressure, and flow condi‐ tions. However, these results certainly demonstrate the power of laser‐based techniques and the three main advantages offered in this particular application. First, this is a direct meas‐ urement of steam quality without any extraction or conditioning ofthe sample. The only inputs required in the method are temperature, pressure, and other standard operating parameters of the steam turbine. Second, the steam quality can be measured over a wide range (100–10%). This is a unique capability over conventional steam calorimeters which have a much smaller range of a (100–80%). Last but not the least, the measurement is real‐time and can be used for better process optimization and steam turbine prognostics.

**Figure 5.** (a) Variation of steam quality with water spray in the in the steam flow pipe and comparison with calculation using mass flow rate. (b) Variation of steam quality with water spray in the steam flow pipe and comparison with calculation using mass flow rate at a different temperature, pressure, and flow rate condition. Figures taken with per‐ mission from IEEE (From our published paper).

#### **4.2. Ammonia slip sensor for gas turbine applications**

This subsection is a summarized excerpt of the work published in reference [14]. Gaseous emissions, such as NOx, SOx, and CO are strictly regulated by the environmental protection agency (EPA) in the United States and similar agencies around the world. To minimize NOx emissions, a selective catalytic reduction (SCR) unit is commonly introduced in the exhaust gas path [52]. The gas temperatures in the harsh SCR exhaust environment are typically of the order of 250–380°C. The functioning of an SCR unit includes injection of ammonia (NH3) gas to cause chemical reactions leading to reduction of NOx into N2 and H2O. This process is depicted in **Figure 6**.

**Figure 6.** Schematic showing the function of a selective catalytic reduction unit in a power plant. (Figure taken with permission from Sharma et al. [14]).

The equations embedded in **Figure 6** suggest that the optimal performance of an SCR would require ammonia injection amount to vary with the NOx generated. Too much ammonia injection leads to incomplete reaction and too little leads to residual ammonia (ammonia slip)

in the exhaust. The amount of ammonia injected is typically controlled by monitoring and minimizing the ammonia slip. The conventional sampling‐based continuous emissions monitoring system (CEMS) [15] is the state‐of‐the art for this application and is often be‐ lieved to be suboptimal due to sampling‐related measurement lags of up to 2–3 min. The most efficient and desired way of controlling the SCR is to directly measure the ammonia slip in the harsh SCR environment (without sampling). An example of how a TDLAS‐based sensor can address this highly challenging real‐world problem follows here.

The TDLAS sensor discussed here is designed for direct operation in an SCR exhaust so that measurement lag time is minimized. The sensor is based on a free space line‐of‐sight meth‐ odology employing a laser transmitter (pitch) to emit the laser radiation and a receiver (catch) to collect the radiation after passage through the SCR exhaust. A diode laser around 1.5 μm is chosen to sense ammonia. Details on spectral analysis and line selection have been present‐ ed in the Section 3.1. In a combined cycle power plant, the SCR exhaust is usually into a heat recovery steam generator (HRSG). The environment inside a HRSG is fairly harsh due to high temperatures, engine vibrations, and floating dust or impurities. The challenges around beam steering caused by temperature gradients and transmission losses due to dust particles are addressed using the 2f/1f WMS methodology described in the Section 2.2. A major challenge specific to this application is the thermally induced misalignment of the beam since the typical line‐of‐sight path length in a HRSG can be up to 10 m. Therefore, maintaining alignment through all engine‐operating conditions is critical toward achieving long‐term reliability.

**Figure 5.** (a) Variation of steam quality with water spray in the in the steam flow pipe and comparison with calculation using mass flow rate. (b) Variation of steam quality with water spray in the steam flow pipe and comparison with calculation using mass flow rate at a different temperature, pressure, and flow rate condition. Figures taken with per‐

This subsection is a summarized excerpt of the work published in reference [14]. Gaseous emissions, such as NOx, SOx, and CO are strictly regulated by the environmental protection agency (EPA) in the United States and similar agencies around the world. To minimize NOx emissions, a selective catalytic reduction (SCR) unit is commonly introduced in the exhaust gas path [52]. The gas temperatures in the harsh SCR exhaust environment are typically of the order of 250–380°C. The functioning of an SCR unit includes injection of ammonia (NH3) gas to cause chemical reactions leading to reduction of NOx into N2 and H2O. This process is

**Figure 6.** Schematic showing the function of a selective catalytic reduction unit in a power plant. (Figure taken with

The equations embedded in **Figure 6** suggest that the optimal performance of an SCR would require ammonia injection amount to vary with the NOx generated. Too much ammonia injection leads to incomplete reaction and too little leads to residual ammonia (ammonia slip)

mission from IEEE (From our published paper).

depicted in **Figure 6**.

406 408High Energy and Short Pulse Lasers

permission from Sharma et al. [14]).

**4.2. Ammonia slip sensor for gas turbine applications**

**Figure 7.** Schematic showing the hybrid ammonia slip sensor with alignment control and implementation in a HRSG. Figure taken with permission from SPIE (From our published paper).

The hybrid sensor, shown in **Figure 7**, is developed to address the misalignment challenge. A 633 nm fixed wavelength laser is multiplexed with the 1.5 μm ammonia spectroscopy laser at the transmitter (launcher) and demultiplexed at the receiver by using appropriate dichro‐ ic mirrors. The 633 nm laser wavelength is chosen because of two reasons: no absorption by exhaust gases and visibility to the human eye (helps in initial alignment of the system). As shown, the launcher optical assembly is mounted on a motion control stage. The demulti‐ plexed red laseris made incident on a quadrant photodiode which serves as a position sensitive detector. The error signal generated by the quadrant photodiode is used as a feedback signal to actively align the motion control stage to keep the red beam centered on the quadrant photodiode. As a consequence, the ammonia spectroscopy beam also remains centered on the 1.5 μm photodiode. This leads to overcoming of measurement errors induced by thermal misalignment and enables long‐term reliable operation of the sensor.

Experimental results demonstrating the performance and value of the alignment control system are shown in **Figure 8**. **Figure 8a** shows the scenario with no vibrations/misalign‐ ments. As expected, the 2f/1f signal stability is the same with the alignment control system on or off. This is a key check to ensure that the alignment control system does not introduce any artifacts or errors by itself. **Figure 8b** shows the scenario under induced vibrations/misalign‐ ments. With the alignment control off, the total normalized transmitted power drops close to zero in a matter of seconds, hence indicating a severe misalignment of the system. However, with the alignment control on, the system is able to actively maintain alignment over time. About 20–25% powerfluctuations are still observed which are experimentally found to be well within the correction capability of the 2f/1f WMS technique. For more details on the tests demonstrating performance ofthe hybrid sensor, the readeris advised to referto reference [14].

**Figure 8.** Experimental results showing the performance of the alignment control system. (a) 2f/1f WMS signals (with and without alignment control) as a function of time under no vibrations. (b) Transmitted spectroscopy laser power (with and without alignment control) as a function of time under induced vibrations. Figure taken with permission from SPIE (From our published paper).

#### **5. Summary**

In summary, this chapter presented the value and potential benefits of diode laser‐based industrial harsh environment sensors. The chapter started with an overview of the industrial Internet and discussed the importance of sensors toward achieving enhanced output from industrial assets such as gas turbines, aircraft engines, and turbomachinery equipment . The discussion showed that real‐time decision making through online sensors (that monitor the desired machine parameters) can enable optimized operation, performance enhancement, and extension of asset life. Subsequently, optical harsh environment sensors were introduced and the capabilities of diode laser‐based techniques, that is TDLAS and QCLAS, were discussed.

The discussion was kept focused on these two most promising techniques (compared to other optical techniques) to remain within the scope of this chapter. The common implementation of TDLAS/QCLAS methodologies, that is direct absorption spectroscopy (DAS) and wave‐ length modulation spectroscopy (WMS), were discussed from an applied experimental perspective. The intention was to equip the reader with just the right level of information required to set up and carry out application‐specific experiments. Appropriate references were provided for readers who wish to go deeper into these techniques. Next, the design philoso‐ phy and methodology behind a TDLAS sensor were discussed. Laser wavelength selection was presented as a crucial step where careful consideration to process temperature, pressure, and background gas interference needs to be given. Optomechanical configurations, namely line of sight, standoff, and extractive sampling, were then discussed followed by instrumen‐ tation and data acquisition basics. Finally, the power of diode lasertechnology in tackling real‐ world challenges was discussed using real‐world examples. An in situ sensorfor measurement of steam quality in the hot and harsh low pressure (LP) section of a steam turbine was presented. A solution based on two diode lasers, each targeted toward water in liquid phase and vapor phase, respectively, was discussed. Last but not the least, the chapter concluded with the discussion on a hybrid in situ ammonia slip sensor for power plant SCR control applications. Anovel alignment control methodology in combination with a 2f/1fWMS scheme was shown to be an effective tool toward measuring gas concentrations in hot and harsh vibrating environments.

The field of optical and diode laser‐based harsh environment sensing is rapidly evolving and is currently an area of active research in academia as well as industry. The various aspects of this technology space are being studied in details by different groups around the world and this chapter has, by no means, touched upon all of them. However, the authors do hope that this chapter has motivated the reader to think of new ideas and concepts that can advance the state‐of‐the art and application areas in this space and help the industrial community realize the full potential of their valuable assets.
