White Paper. Gas Analysis Made Easy Plug and Play Tunable Diode Lasers INGOLD

Transcription

1 INGOLD Leading Process Analytics White Paper Gas Analysis Made Easy Plug and Play Tunable Diode Lasers When it comes to process safety for protecting people, the environment and assets, only the best equipment will do. This is particularly true for gas measurement in industrial processes. Faced with a wide range of measurement options, users must carefully select the right technology for each measurement task. Latest developments in the field of Tunable Diode Lasers make this technology particularly attractive for achieving efficient and reliable measurement of single gas species. Extractive oxygen measurement systems often play a major role in keeping manufacturing processes safe and maintaining inert conditions in any given situation. As a result, they must stand up to the toughest conditions and as a consequence, significant maintenance is required to keep their performance at an acceptable level. Typically, the total lifetime maintenance costs for an extractive oxygen analyzer with sampling and conditioning system can add up to twice the purchasing, engineering and installation costs incurred during the initial phase. This not only burdens maintenance budgets, it also jeopardizes the overall process yield since paramagnetic systems can fail abruptly if moisture and dust enter the measurement cell. These significant drawbacks can be overcome by using Tunable Diode Lasers (TDL) that are able to measure in situ, with high accuracy and reliability, and with virtually no maintenance.

2 Gas Analysis Made Easy Laser technology for interference-free analysis Oxygen analyzers that are able to provide faster measurement and more accurate results are always going to be of interest to the chemical industry. If they also have a significantly lower cost of ownership compared with other systems, it adds up to a very attractive value proposition. In narrow-bandwidth laser spectroscopy, the absorption of light is measured at discrete lines using finely tuned VCSEL or DFB lasers routinely used in the communications industry. Such laser diodes can be tuned to emit a very specific wavelength that matches the part of the spectrum where single absorption lines for a particular gas species to be detected are available. By controlling the laser temperature as well as the feed current, a narrow scan of the selected absorption lines can be performed in order to determine the surface area of the lines and hence the concentration of the gas species. Using absorption spectroscopy, gases frequently measured in industrial processes such as oxygen can be measured in situ, directly in the gas stream and bypassing any sampling and conditioning. This is achievable up to temperatures of 1500 C or 10 bar and excludes any possible interference from background gases. Because absorption is attenuated due to the presence of dust in the gas stream, a reduction of signal intensity can be expected. However, most TDLs on the market are able to cope with 90 % attenuated signals. Simplified installation and commissioning This technology has been known for over two decades, and throughout this period has continuously been refined to improve performance and drive down costs. However, users of TDLs must also be aware of the effort and care needed in order to make these analyzers perform the way they should, right from day one. Areas where users have frequently reported difficulties are: Selection of the measurement point for temperature and pressure inputs: Absorption lines change shape with temperature and pressure (see Fig. 1), so in order to calculate a correct concentration value, temperature and pressure of the gas stream must be measured. This is usually done by external sensors mounted in the pipe. If the value provided by the sensors is not representative of the gas to be measured, a considerable error results. As a rule of thumb, a 50 mbar deviation results in a 1 % deviation of the air concentration value. Selection of the right purging speed: TDL is a non-contacting measurement method; therefore, the laser diode and detector must be separated from the process by windows. In a typical cross-stack TDL setup, separate purging lines for the sender and the receiver unit on both sides of the stack must be pulled. The purge gas, usually nitrogen, flushes the windows process side to keep them free of detrimental dust sedimentation that could rapidly decrease signal intensity. What is more, the purging speed needs to be carefully selected, as a build-up of nitrogen shortens the measuring distance of the laser and results in measurement error. Alignment of flanges for sender and receiver units: Although the laser beam is by no means as collimated (parallel) as a laser pointer for presentations, both units need to be aligned within 1 2 for the beam to hit the detector. So flanges must be welded on both sides of the pipe with high accuracy for basic alignment, and the fine adjustment is then done by positioning screws on the sender or receiver unit. It is not uncommon that this alignment procedure needs to be repeated when the hot process is running, to account for different thermal expansions on both sides of the pipe or conduit. Temperature Pressure 23 C / 73.4 F Path Length 300 C / 572 F 5 bar(a) ) 72.5 psi(a) 1 bar(a) 14.5 psi(a) 2 m / 6.56 ft 1m / 3.28 ft Fig 1: Influence of temperature, pressure and path length on absorption lines 2

3 Gas Analysis Made Easy In-situ TDL in probe design Cross-stack design for in-situ TDLs is frequent, but not the rule. A valid option to overcome some of the installation hazards are probe-style TDLs. A probe is a part of the sensor that protrudes into the process gas stream. In a probe TDL, the laser diode and the detector sit on the same side of the pipe. The outgoing beam is reflected back to the detector using a corner cube (see Fig. 2 below). A corner cube has the ability to reflect back the laser beam parallel to the outgoing beam, at any orientation of the corner cube. So the optical path is aligned by factory settings. This TDL probe design offers significant advantages over crossstack TDLs: Alignment of the flanges is obsolete, since the whole optical path is geometrically set by the probe itself, offering plug-and-play installation. The spectrophotometer is detachable from the probe, that also serves as a process interface sealing from the gas stream. So maintenance to the analyzer is possible without process interruption. The optical path is folded, which means that for the same insertion length, the beam path is doubled, thereby doubling the accuracy of the sensor. Consequently, good detection limits can be reached even in small pipe diameters, which greatly improves the ability to measure gas concentration where the process requires it, not just where the cross-stack TDL can be accommodated. Additionally, advances in processing power have allowed TDLs to become more compact, so that purged enclosures for use in hazardous areas can be replaced by explosion-proof ones that comply with FM and ATEX regulations. This also means that for hazardous area compliance, it is no longer necessary to use long and costly fiber connections between process gas-wetted elements and the analyzer electronics. This can become a significant advantage when selecting the installation point in the process, since the footprint of probe TDLs is minimal. Advanced signal processing State-of-the-art TDLs for process and safety applications do not only make use of the latest advances in stable and long-lasting diode laser sources; they also profit from advanced signal processing algorithms that are able to cope with a wider range of process conditions and can deliver stable and reproducible concentration values. One key recent development concerns the issue of line locking, a general concern for TDLs where it must be assured that the wavelength range scanned by the laser diode exactly matches the wavelength range where the expected absorption peak is present, even after long periods of time where the TDL is powered but does not measure any gas concentration. With the SpectraID technology for oxygen measurement, three consecutive absorption peaks situated next to each other are analyzed in height, relative position and area. These results are then compared with a physical model of the absorption lines. If there is a match between the two sets of data then there is a perfect DNA match and it can be concluded that the observed absorption peaks are fully identified (see next page, Fig. 3). Bright future Tunable Diode Lasers are generating major interest due to their incomparable benefits for measurement reliability: absence of interference, long-term signal stability, and virtually no maintenance. From the supplier side, signs abound of a wealth of innovations that will further improve the user s experience: easy installation and commissioning, and additional signal robustness are a few examples of recent improvements that are already available. Looking at near-future developments, certainly the commercial availability of laser diodes for the micrometer range will open new horizons to more gas species to be measured, making TDLs a true platform for gas analytics. Fig 2: Probe-style GPro 500 TDL with folded path configuration and corner cube (left) 3

4 Gas Analysis Made Easy Absorption 1. Measured spectrum at known pressure and temperature 760 nm Wavelength 2. Full spectral database for the same conditions 761 nm Absorption 760 nm Wavelength 761 nm 3. Exact match of spectral data for the determination of the concentration. 5.5 % O 2 Fig 3: Three consecutive absorption peaks situated next to each other are analyzed in height, relative position and area For more information, visit: 4www.mt.com/o2-gas Mettler-Toledo AG Process Analytics Im Hackacker 15 CH Urdorf Switzerland For more information 02 / 2012

5 INGOLD Leading Process Analytics White Paper In chemical plants, petrochemical plants, and refineries tunable diode lasers (TDLs) are becoming an increasingly common sight. Their high reliability and low maintenance has made them the gas analyzer technology of choice for many companies. However, installation locations and conditions encountered in some processes have limited their application range. A portfolio of folded-path TDLs with a unique range of adaptions has opened the door to measurement opportunities previously considered impossible. TDLs for All Your Processes Folded-Path Gas Analyzers Introduction The rapid rise of TDL analyzers over recent years has led to them becoming established as a core measurement technology within the portfolio of gas analysis techniques available today. The ability of TDLs to interface directly to the process, which eliminates the need for costly and high-maintenance sample handling systems, gives them an inherent rapid speed of response. This makes them ideal for real-time dynamic measurement of process conditions. Conversely, one aspect of TDLs has not advanced at the same pace as the measurement technology itself: the process interface available for cross-stack, in situ units. Installing an optical, cross-stack TDL directly in a process pipe or vessel creates some installation and operational challenges and limitations which need to be considered carefully before considering their deployment.

6 Figure 1: Classic cross-stack in situ TDL alignment. and measurement limitations on small pipes but, due to their large internal volumes, also exacerbate the stability of alignment, and provision and consumption of purge gas required to keep the analyzer s optics clean. Optical alignment Even if space limitations are not a concern, alignment of the transmitter and receiver units across the pipe is always a consideration. As Figure 1 illustrates, a cross-stack TDL requires careful alignment, which means mating flanges have to be welded onto the pipe prior to installation, adding cost and complexity to the procedure. Figure 2: Difficulties of cross-stack measurement on flexible thin walls and hot processes. This paper discusses these considerations, and details new innovative process adaptions that overcome the limitations to allow the operation of TDL analyzers in locations and applications previously considered impractical, if not impossible. Installation point selection and unit size The first consideration when planning a cross-stack/pipe TDL is the installation point. The decision has to be based on executing the measurement at the point in the process where the most pertinent analysis data can be collected. However, this can create the first challenge if there are space constraints or if the diameter of the process pipe is small. Optical restrictions mean that an elongated housing is needed to provide the required focal length between the laser and the receiver. These long optical housings not only create installation To aid alignment of the transmitter and receiver units, various alignment mechanisms are available, ranging from simple, large, compression O-rings to more complex, flexible, metal sealing designs. Care has to be taken to ensure that satisfactory alignment has been achieved, and that the integrity of the process line has not been compromised through introducing leak paths during the alignment process which could allow the escape of process gases into the atmosphere. Physical mass An issue that is particularly relevant when considering installing a cross-stack TDL relates to its size. Firstly, the optical alignment can degrade over time as the alignment mechanism sags due to the weight of the transmitter and receiver body together with the long optical housings (Figure 2). Additionally, if the vessel wall is thin it is essential to add braces to provide additional stability (Figure 3). 2

7 Even when these considerations have been thoroughly accounted for, alignment can still be compromised if the process temperature varies greatly, for example, during start-up or when using the analyzer on a batch process where there is a significant temperature ramp profile. This is because the walls of the process vessel will flex as they heat up, and although this may not be visibly noticeable it can lead to the laser becoming misaligned to the receiver once the process is running, necessitating expensive and time-consuming realignment. Figure 3: Bracing the process vessel wall to stabilize alignment and diverging the beam to ensure adequate portion of the signal is received. In an attempt to mitigate this effect, the laser beam profile can be altered to create a divergent optical spread (Figure 3). While this can help to ensure some signal always reaches the detector, the received signal intensity is inevitably reduced. Figure 4: Bypass installation of a cross-stack TDL. Purge gas The next major consideration is the provision of purge gas to protect the analyzer s optical windows from particulates and/or condensation entrained in the process gas stream. The majority of cross-stack TDLs require an optical purge, and due to the diameter of most TDL optical housings there is a large purge volume to fill and a sizable optical surface area to be kept clean. This leads to significant purge flow requirements, sometimes as much as 50 liters/min (1.77 cuft/min) for each side. Such a high level of consumption obviously brings cost implications while, additionally, it introduces substantial quantities of diluent gas into the process stream which may create process quality or other problems downstream. Small pipes For many processes it would be ideal if it were possible to easily employ a TDL analyzer directly across a small bore process pipe. This is a challenge for the majority of in situ TDLs due to their size, weight, sensitivity, and optical design limitations. Even the best instruments available are usually limited to a minimum pipe diameter of 12 (DIN 300). This restriction often means that a suitable expansion pipe section has to be installed, which brings its own concerns (pressure reduction and reduced velocity), or a slipstream or bypass installation must be fitted (Figure 4). 3

8 Although the bypass approach can enable a crossstack TDL to be installed Cold purge gas on a small pipe, it is not an ideal solution. Firstly, in such an arrangement the optical purge on one Transmitter side is working in the direction of process flow, while the other is working against the flow. This can lead to instability of the optical path length which in turn will result in measurement error. Secondly, the purge gas will have a pronounced dilution effect on the process gas as it passes through the bypass arm. This inevitably results in a measurement that is not truly representative of the process sample. In addition, this approach suffers from other drawbacks such as high optical purge gas consumption and process flow control considerations, especially as the valves can introduce undesirable pressure drops in the line. Reflection Diffraction Figure 5: How beam steering affects a TDL analyzer. Steered beam Receiver Beam without steering random fluctuations of the laser beam at the receiver and therefore noise on the signal. In addition, the multiple signal paths further contribute to increased signal noise and measurement instability. So, in addition to the costs involved in supplying the purge gas to the measurement location and the high purge consumption, the purge gas can, under some circumstances, cause degradation of the received signal. Beam steering The final consideration is more subtle but nonetheless is an important phenomenon that is relevant to TDL technology. When a laser beam passes through gases at various temperatures and therefore different densities and indices of refraction, the beam will be diffracted at the interface where the density changes. When using a cross-stack TDL to analyze a hot gas stream this is exactly what occurs, since typically there will be a non-homogeneous temperature distribution and in addition, a high flow rate of cold purge gas at each end of the optical path. This diffraction of the laser light is known as beam steering and contributes to signal noise. It affects measurement stability of the analyzer as well as potentially adding complications to the alignment process. As can be seen in Figure 5 the laser and detector have been prealigned across the process pipe (the dotted line). Once the hot process gas is flowing, due to the fronts created by the different indices of refraction, the laser beam will be diffracted and some of the beam may also be reflected. This leads not only to loss of received energy at the detector, but importantly, laser energy will arrive at the receiver from various optical paths and some rays may even miss the detector entirely. The cumulative effect is In general, the beam steering effect increases with temperature and pressure in the process and also with the length of the optical path through the hot gas. Even if there is no cold purge gas, beam steering can still cause a significant decrease in measurement quality. In many cases, the benefit of a long optical path length is mitigated by the signal noise introduced by beam steering. Therefore, the optical path length should be carefully chosen depending on the process conditions. As has been outlined above, there exists significant installation constraints that must be understood when planning to implement cross-stack TDL analyzers. This raises the question: Can these constraints be removed so that it is no longer necessary to adapt the process or plant infrastructure to fit the analyzer, but instead to make the analyzer fit the plant? Make the analyzer fit the process Where can choices be made to overcome the limitations outlined above? As a general rule a reduction in the weight of the TDL can go a long way to eliminating concerns regarding alignment stability and the use of TDLs in tight spaces. Some lighter crossstack analyzers are available, but they still require alignment, which can become more difficult if the optics and beam diameter 4

9 have also been reduced. There also still needs to be opposing flanges fitted across the pipe and significant volumes of purge gas. When TDLs are being considered it is often assumed that they have to be crossstack devices. But there is an alternative. For the majority of process applications a folded-path TDL (where the laser beam from the sensor head is reflected back to a receiver also in the sensor head) offers a host of advantages: Transmitter and receiver in a single unit and no need for an expensive interconnecting cable. Usually single flange installation No tricky alignment across the pipe or vessel Order of magnitude reduction in purge gas requirement Small size, so easy to install in tight spaces Greater accuracy (as the laser beam passes through the sample twice) Lightweight design removes stress on flanges and seals. Figure 6: Example of a folded-path TDL mounted across a 4" (DN100) pipe. Figure 7: Purge gas dominates optical path with a purge-type probe, but non-purged probe offers a solution. Folded-path TDLs (Figure 6) address many of the problems of cross-stack analyzers. The advantages they offer have significantly opened up measurement opportunities in situations that have previously proved difficult or impossible to overcome. Such TDLs may still require purge gas to protect the optics, but due to the smaller diameter of the in situ probe and low internal volumes the purge requirement is significantly lower, sometimes as much as an order of magnitude lower. For gas streams containing entrained particles or where there is potential for condensation, purge will remain a necessity, at least for the foreseeable future, but not all processes fall into this category, as discussed below. 5

10 Headspace monitoring using a TDL When considering a TDL for headspace monitoring, the matter of purging is critical. This is because the purge gas will dominate the optical path of the TDL (Figure 7) as the sample is largely static and there will typically be insignificant circulation or velocity in the headspace to displace the purge gas. This would be a hazardous situation, as it would create a non-fail-safe condition; since the analyzer will always show an erroneously low target gas value. In order to provide a secure and effective measurement, the use of purge gas is, therefore, not possible. Since many headspace and inertization applications are free of particulates, not using purge gas can be viable. But, for cross-stack analyzers two ports are still required. This eliminates simple attachment to the tank and means elaborate bypass or extractive systems are necessary. A better solution is a folded-path TDL that does not require purging and that can interface through a single port into the headspace itself. The separation between the optical windows of the device fully defines the optical path, and as the laser beam passes twice through the gas, good sensitivity can be achieved with a compact probe. As purge gas is not required the entirety of gas within the optical path will be at the same temperature and density, therefore there will be no beam steering effect. This can significantly improve both signal stability and sensitivity. Hot and dusty applications Another application that poses issues for deploying TDLs is hot and dusty processes. If purge gas entering the gas matrix is not a problem then it can be used to protect the optics on either a cross-stack or purged folded-path TDL. But the combination of cold purge gas and hot process gas raises the problem of beam steering. The ideal solution would be to remove the purge requirement if possible (with the added benefit of simplified installation and reduced running costs). However, removing the purge means another form of protection for the optical windows is necessary. The answer here is use of a non-purged, folded-path TDL with the addition of a sintered metal filter and baffle (Figure 8). Figure 8: Example of a TDL with a non-purged (NP) process adaptor and particulate filter. Figure 9: TDL with direct process pipe (wafer cell) adaptor. Figure 10: Extractive process adaptor. This unique adaptor allows a TDL to be installed on pipes down to 2" (DIN 50) and offers no obstruction to the process flow, which brings the additional benefits of compatibility in high velocity operation (> 25 m/s) and immunity to vibration. Process adaption for extractive systems There are occasions when there might be a fully-serviceable sample handling system, perhaps one previously used with a paramagnetic or NDIR analyzer, that a site would like to continue to use with a TDL installation. There are many reasons that this might be an appealing option. It allows a simple means of calibration/validation (the process might be challenging with high levels of condensates or extreme particulate loads), where use of a sample conditioning system offers the most appropriate solution. TDL on pipes down to 2" (DIN 50) In the past, a typical no-go location for TDLs has been in situations where there is a need to interface directly into a small diameter process pipe. As has been discussed, cross-stack analyzers require complex and expensive bypass configurations, and even folded-path analyzers are typically limited to a minimum diameter line of 4 (DIN 100). An innovative in-line process adaption: the wafer cell (Figure 9), provides the solution. Previously, adding a TDL analyzer in these circumstances has usually meant either; installing a specialized, fixed, extractivestyle TDL; or adding a sample cell to which the transmitter and receiver housings of a cross-stack analyzer can be attached. This is a viable solution, but will often mean a large panel is required to hold the sample cell and analyzer hardware and can be a significant problem if wall space is at a premium. In another scenario a test of the performance of a folded-path TDL might be required before committing to an in situ installation. 6

11 Cross-stack Probe-type Non-purge Wafer cell Extractive cell probe-type Alignment required Yes No No No No Purge gas required Yes (large Yes (small No Yes (small No quantity) quantity) quantity) Can be used on small pipes Typically, no Yes 4" Yes 4" Yes 2" N / A (DIN 100) (DIN 100) (DIN 50) Beam steering Yes. Yes. No Yes. No Can be Minimal due Minimal due significant if to short path to short path path length is length and length and long optical design optical design Table 1: Comparison table of cross-stack TDL versus folded-path TDL with probe and process adaptions. An elegant approach in these circumstances is an analysis device that can be easily adapted between an extractive or in situ interface, whether it is a purged, or non-purged probe, or an in-line wafer cell. Such an extractive process adaptor (Figure 10) attaches directly to the TDL analyzer to create a stable, fixed, and pre-aligned optical cavity. The analyzer can be operated continuously as an extractive analyzer, or modified later simply by changing the process adaptor to offer the most appropriate interface for the final in situ location. technology to finally be taken right to the heart of the process plant, without all of the restrictions that this has previously meant. By recognizing these limitations and re-imagining a better solution, it is now possible for a compact, lightweight TDL analyzer to be utilized across a vast range of installation locations with confidence and without compromise. 4www.mt.com/TDL Conclusion Throughout this paper emphasis has been placed on the benefits of TDL measurement, benefits that are not always matched in practice due to the limitations of the process interface of typical cross-stack, in situ TDLs. These limitations create challenges and often lead to compromises regarding measurement stability and integrity in many potential TDL applications. A re-thinking of these challenges has led to the development of a series of new innovative process adaptions for folded-path TDLs. These adaptions allow this powerful optical measurement Mettler-Toledo AG Process Analytics Im Hackacker 15 CH-8902 Urdorf Switzerland For more information 05/2014