Dr. Christopher J. Myatt and Nicholas G. Traggis, Precision Photonics Corp.
Monitoring the wavelength reflected from a fiber Bragg grating can be achieved using tunable laser technology combined with absolute wavelength referencing. Laser sources are preferred for applications such as overseeing the operating conditions of oil wells, because their spectral power density yields high signal-to-noise levels.
This method works by sweeping a narrowband tunable laser source over the spectral range, including the fiber Bragg grating’s spectral peak. The laser wavelength measurement module records the wavelength of the scanning laser, while the power meter records a peak in the reflected intensity. One advantage to this approach is that laser choices are abundant as a result of the big telecommunications buildup.
Fast-sweeping test-and-measurement lasers have been developed, and widely tunable telecom transmission lasers have finally reached deployment. Any of these devices is adequate, although to satisfy the stability requirements, control architectures must be considered for each configuration. Tuning speed and characteristics, the maturity and ruggedness of the technology, the supply chain and output power all factor into a design choice.
Many of the interrogator solutions being developed and deployed make use of mature test-and-measurement laser technology. These units are typically based on an external-cavity laser design consisting of a laser diode, an external dispersive device such as a grating, a mirror and a tuning actuator. The tuning actuator is often a motor that provides ruggedness and a large range of motion. The motor technology is well-developed and allows very smooth and fast tuning, reducing the burden on the wavelength-characterization system. These units also allow wide tuning ranges of 100 nm or more with very little power or frequency fluctuation (mode hops).
If less tuning range is required, the technology developed for telecom tunable transmission lasers is available. External-cavity modules that have a Littman-Metcalf microelectromechanical system (MEMS) tuning mechanism offer many of the benefits of the test-and-measurement units, with potential cost benefits in deployment.
Tunable laser modules based on vertical-cavity surface-emitting laser technology, with a MEMS cantilevered cavity mirror, offer a smooth tuning profile but are hampered by lower power and a broader linewidth.
A tunable laser spectrum analyzer system can achieve precision of 0.01 pm. This results in the ability to detect nanostrain or a 10-ppb shift of a spectral feature.
Another option is a thermally tuned laser module. Examples include single and array-based distributed feedback diode lasers and some external-cavity laser types. Although acceptable for sparing applications in telecommunications networks, these are not a suitable solution because they’re slow when tuning across the desired band.
The last telecom-influenced solution is an electronically tunable distributed Bragg reflector or sampled-grating distributed Bragg reflector structured diode laser. This is a multiple-current laser with a separate gain, wavelength tuning and phase current. In principle, these electronically tuned units are rugged and fast tuning over a wide range, but their construction requires complicated lithography, and they are being offered by only one supplier.
Systems based on telecom tunable lasers offer the promise of lower cost, and those that achieve broad market acceptance will be available at the lowest cost.
Wavelength calibration
The goal of interrogation equipment for fiber Bragg grating sensors is to precisely characterize the peak reflectivity of the grating. The interrogation equipment sends light out to the fiber Bragg grating and analyzes the reflected light as the wavelength of the source is changed. To achieve wide dynamic range in sensing the environmental variable requires high-precision wavelength calibration. Such precision allows small changes in the process conditions to be detected on large background levels. For example, a 3.25-psi drop in pressure in an oil well, where static pressures reach upwards of 10,000 psi, can indicate a significant change of flow in the crude oil stream.1 In terms of wavelength measurements, resolving such pressure shifts requires precision at the level of 0.0001 nm (0.1 pm) over 10 nm. There are similar requirements in other process control applications, such as measuring temperature and pressure in refinery reactor vessels, electric-power-generating turbines and jet engines. To maintain such precision, the wavelength calibration equipment must seamlessly integrate with the laser source.
Calibrating the wavelength of a widely tunable laser source presents several technical hurdles. First, the calibration must work over a large tuning range (typically 40 to 100 nm), so that many sensors can be used. Second, it must be highly accurate and precise in a short measurement time so that fast update rates are possible. For instance, today’s tunable lasers have scan speeds in excess of 100 nm/s, leading to update times of less than 1 second. Finally, the wavelength calibration must be perfectly synchronized with the sensor response data for accurate measurement of peak shifts. Although traditional wavelength meters are highly accurate over a large wavelength range, they have slow measurement times and are not suitable for measuring fast-tuning lasers.
Several products have recently been introduced to specifically address wavelength calibration of fasttuning lasers, enabling high-accuracy calibrations at scan speeds exceeding 100 nm/s over a 100-nm tuning range. These systems often involve comparing the laser light with an interferometric reference and using stored calibration data to obtain an accurate readout. For telecom applications, this is adequate and reaches precision at the few picometers level. Pushing beyond this level of stability requires real-time calibration to an absolute reference such as a gas cell.
One measure of the stability of a system is captured in a statistic called the Allan deviation, the standard deviation of a measurement as a function of the number of measurements averaged together.2 In other words, a plot of the Allan deviation indicates how measurement precision is improved by “averaging down.” The ability to improve a measurement through averaging is also called stability. Without stability, averaging over time does not improve the precision of the measurement. Over very long times (hours, months, years), measurement stability is critical for process monitoring. In the oil field example, a reservoir engineer needs to compare current pressure and temperature readings with data accumulated over the life of the well to maximize oil extraction.
A plot of the Allan deviation shows the stability that can be achieved with precision laser interrogators (see figure). In a single scan, the system tested can achieve approximately 0.07-pm precision. With averages of up to 30 successive measurements, the precision can be improved to 12 fm (0.012 pm). With a resultant resolution of one part in 108, this is one of the most precise measurement systems on the market.
The data indicates that further averaging does not improve the precision, showing that the stability limit of this equipment has been reached. Clearly, this system can deliver the subpicometer precision and stability required to detect subtle shifts of temperature and pressure in an oil reservoir as it is drained.
As demonstrated in the data shown and in research results over many years, optical fiber sensing can now meet the requirements for precision process measurement and control.
References
1. W.N. Barkhouse (2003). Paper 15075. Offshore Technology Conference 2003 Proceedings.
2. D.W. Allan (February 1966). Statistics of Atomic Frequency Standards. Proceedings IEEE 54, p. 221.
Meet the authors
Christopher J. Myatt is the founder and CEO of Precision Photonics Corp. in Boulder, Colo.
Nicholas G. Traggis is director of new product development at the company