Grantee Research Project Results
Final Report: Hand-Held Sensor for Remotely Mapping Carbon Dioxide Pollution Sources
EPA Contract Number: EPD10056Title: Hand-Held Sensor for Remotely Mapping Carbon Dioxide Pollution Sources
Investigators: Roos, Peter A
Small Business: Bridger Photonics, Inc.
EPA Contact: Richards, April
Phase: II
Project Period: May 1, 2010 through April 30, 2012
Project Amount: $225,000
RFA: Small Business Innovation Research (SBIR) - Phase II (2010) RFA Text | Recipients Lists
Research Category: Small Business Innovation Research (SBIR) , SBIR - Air Pollution
Description:
In 2007, the U.S. Supreme Court ruled that carbon dioxide (CO2) is a pollutant under the federal Clean Air Act. The ruling allows the EPA to regulate CO2 emissions. Such regulation will entail monitoring a wide variety of pollution sources, including automobile exhaust systems, industrial emission sources and carbon sequestration sites. With presently available technologies, EPA personnel will need to perform onsite scans of possible pollution locations by tediously sampling emitted gases with point-source gas-intake measurement devices. This makes it difficult or impossible for EPA personnel to identify or quantify critical CO2 pollution sources, such as many smokestacks/vents or unknown leaks in large search areas. No commercial technology currently exists that can remotely measure and pinpoint (to within a few meters) the location of elevated CO2 concentrations.
Through an EPA SBIR Phase II effort (including a Commercialization Supplement), Bridger Photonics, Inc. (Bridger) set out to fill this technology and market need with an innovative laser-based remote gas sensor. Bridger's effort focused on two overall goals:
- Demonstrate the critical CO2 gas sensing capabilities of this technology with a breadboard prototype. With the resources available, Bridger designed the effort to provide the groundwork and data needed to establish the prototype viability.
- Develop the compact laser source (i.e., without the gas sensing capabilities) into a robust, marketable product to generate revenue that could be used to help fund the full gas sensor commercialization.
Summary/Accomplishments (Outputs/Outcomes):
Bridger was successful in accomplishing both overall goals of the project. The technical approach for Bridger's range-resolved gas sensor was based on Differential Absorption LIDAR (DIAL) at 2 µm wavelength, where the CO2 absorption strength is ideal for high spatial resolution concentration measurements. However, no suitable laser transmitters exist at this wavelength, so a large portion of the effort was devoted to innovating a 2 µm laser source based on frequency conversion of a diode-pumped Nd:YAG laser. Not only did Bridger achieve a suitable laser transmitter for its gas sensor, but thanks in large part to the EPA's Commercialization Supplement, Bridger was able to develop this technology into a compact and efficient mid-infrared laser product capable of generating up to 3 mJ pulse energies at wavelengths that can be factory selected between 1.4 µm and 4 µm wavelength. As per goal #2 above, Bridger is marketing this product as a stand-alone unit (see Figure 1) and has already sold and delivered nine of these laser systems for a laser ablation application. The revenue from these sales already surpasses the monetary investment made by the EPA in this project. Moreover, the EPA's investment in this project will help meet the EPA's gas sensing needs even beyond CO2 detection and monitoring. For instance, the laser product has the flexibility to operate at the wavelengths necessary to detect methane for the EPA's Natural Gas STAR program.1
Figure 1. Bridger's MIR Series mid-infrared laser generates
up to 3 mJ pulse energy from up to 7 mJ pulses 1064 nm
(before frequency conversion to the infrared).
Bridger had equal success demonstrating the ability to use this laser technology as a transmitter for CO2 detection. The receiver for this system consisted of a 120 mm diameter telescope, which focused the returning 2 µm light onto Bridger's detector. Bridger devoted considerable resources to developing a low noise detector with sufficient bandwidth. In its final design, Bridger demonstrated a bandwidth of 35 MHz and average NEP of 8e-12 W/sqrt(Hz), as described in the full report.
Using the developed transmitter and receiver, Bridger took many gas measurements both with topographical scatterers (where the light scatters off an object to provide large returns, but no range information) and with atmospheric backscatter (where the light scatters only off particles and molecules in the atmosphere to provide a distributed reflector). The three measurements that best characterize the gas sensing results are shown in Figure 2. Figure 2 (left) shows the transmission of Bridger's laser source after it has been passed through a gas cell containing CO2. The dips in transmission as a function of laser wavelength indicate that Bridger's laser light is sufficiently narrowband and emitting at the proper wavelength to be efficiently absorbed by the R branch of the 2 µm CO2 absorption band. The line represents predictions from the HITRAN database. Similarly, Figure 2 (center) shows the ambient atmospheric transmission versus wavelength taken outdoors over a one-way path length of 100 m with a topographical scatterer. Circles in the plot indicate the CO2 absorption lines, while the other absorption lines are due to water vapor. Bridger's DIAL measurement technique relies on measuring the difference in absorption between when the lasers wavelength is tuned on and off an absorption line. Representative results showing atmospheric CO2 sensing as a function of range with no topographical scatterer are shown in Figure 2 (right). From the plot, the breadboard prototype is capable of CO2 concentration measurements with about 100 ppm sensitivity out to beyond 50 meters (10 minute measurement duration). To put this measurement into perspective, typical human breath is 5,000 ppm and industrial emissions are often 30,000 150,000 ppm CO2 concentration. Although Bridger focused on <100 m ranges for onsite plume measurement applications, the sensor can reach longer ranges and can take faster measurements for applications where sensitivity is not as critical.
Figure 2. Left: Transmission through a CO2 gas cell versus wavelength. Center: Ambient atmospheric
transmission using a topographical scatter and a 100 m path length. Right: Range-resolved atmospheric
CO2 measurements.
The potential revenue from the gas sensor after it is commercialized also is promising. The recent U.S. Supreme Court ruling combined with emerging international efforts to reduce CO2 emissions establishes a growing market demand for the hand-held CO2 sensor. Bridger estimates a potential $50M U.S. market for this device, including future carbon sequestration monitoring markets. This technology can be easily transitioned to detect other gases that are of interest to EPA and others.
2. TECHNICAL OBJECTIVE RESULTS HIGHLIGHTS
2.1. Objective 1. Optimize the performance of the laser system
Bridger's CO2 sensor design, based on Differential Absorption LIDAR (DIAL), depends heavily on a high-performance laser transmitter. The transmitter must emit narrowband 2.0 µm wavelength light with sufficient energy, average power, and beam quality to perform DIAL measurements using atmosphere backscatter. Bridger's stated initial design goals were to build a transmitter with 70 percent seeding efficiency, 10 Hz pulse repetition rate and greater than 1 mJ pulse energy. Bridger's transmitter surpassed the initial performance goals with 96 percent seeding efficiency, 300 Hz pulse repetition rate and 500 uJ pulse energy. All tasks for this objective were completed on time and with no significant problems. A few highlights of Bridger's technical advances on optimizing the laser performance are described here.
Pulse Repetition Rate
Bridger's initial stated goal was to achieve a 10 Hz repetition rate for this system. However, higher pulse repetition rates translate into faster LIDAR measurements with better sensitivity, so Bridger took the opportunity to increase the pulse repetition rate beyond its initial goal. The primary inhibiting factor in achieving a higher repetition rate is thermal lensing in the laser crystal. As the laser repetition rate is increased, thermal lensing becomes stronger, and the pulse energy decreases eventually to the point where the laser will not operate.
To achieve high pulse repetition rate, Bridger improved the efficiency of the laser (thereby reducing heat deposition) by increasing the mode overlap of the 808 nm diode stack emission with the resonant spatial mode of the laser. A diagram of the laser components is shown in Figure 3. This was done by collimating the output of the diode using a lenslet array and focusing into the back of the laser crystal. Spatial hole burning problems in the laser crystal were improved by triple pulsing the laser, which was done by operating the diode pump laser continuously until three pulses have fired. This allows for neighboring longitudinal laser modes to lase on successive pulses, thereby reducing the effect of longitudinal spatial hole burning in the gain medium. Through these techniques, Bridger was able to operate the laser with pulse repetition rates of up to 300 Hz (30 times higher than the stated goal). This dramatic improvement in repetition rate far outweighed the fact that Bridger chose to produce 0.5 mJ pulses rather than its desired 1 mJ to avoid component damage. This repetition rate allowed the results shown in Figure 1 and is still limited by thermal lensing in the gain medium.
Figure 3. Left: Image of a pump laser utilizing a robust, adjustment
free monolithic design. Right: 1064 nm pump laser. Nd:YAG laser gain
medium. QS is a passive Cr: YAG q-swtch. LP is a thin film linear
polarizer. and OC is a 70% reflective output coupler.
Seeding Efficiency
The gas sensing results shown in Figure 2 are obtainable only when the mid-infrared spectral emission of the system is narrower than the spectral features of the gas (about 3 GHz, see Figure 2, left and center). During this Phase II effort, Bridger used injection seeding with a diode laser at the signal wavelength to narrow the spectral emission. Bridger's initial goal was to achieve 70 percent of the emission narrowed to the spectral width of the seed laser. Figure 4 shows the output emission as a function of wavelength when the seed laser is carefully matched to the signal mode of the parametric generation process. The blue trace in the figure shows the strong narrowing of the output emission when seeded. Bridger achieved high seeding efficiencies of 96 percent (exceeding its goal of 70%) for pulse energies below 0.5 mJ. The undesired degenerate mode also is shown in the inset before seeding. This mode is suppressed to a negligible level through the seeding process.
Figure 4. Seeded and unseeded 2 µm wavelength light showing
significant suppression of the degenerate mode.
2.2. Objective 2. Construct Optical Receiver
In parallel with the successful laser transmitter development, Bridger also developed a receiver for collecting and detecting the backscattered light from the transmitter. All of the tasks for this objective were completed, and Bridger was even able to accomplish critical modeling that was not originally planned. The detector optimization task required significantly more time and effort than originally anticipated, but in the end Bridger was able to achieve a detector design that served its desired purpose. Highlights from the receiver construction follow.
The most critical components of the receiver were the telescope and detector. Bridger developed a model in MATLAB to optimize the overlap function of the receiver telescope field of view with the transmitted beam. The telescope had a120 mm aperture and a 650 mm focal length. The focused spot size became smaller than the detector for ranges greater than 40 m. As described in the full final report, the model predictions were validated with experimental results.
Once the light was collected, efficient and low noise detection proved to be the most challenging aspect of this effort. Bridger constructed a photodetector using an extended wavelength InGaAs photodiode and a high-speed, low-noise transimpedance amplifier. After a number of iterations, the detector demonstrated a bandwidth of 35 MHz and average NEP of 8e-12 W/sqrt(Hz) over the entire bandwidth. Figure 5 (left) shows the detector NEP as a function of RF frequency before (green trace) and after (blue trace) the last iteration of the detector design. Bridger found that careful design of the detector circuit allowed critical suppression of undesired noise. Also on this contact, Bridger demonstrated custom deconvolution algorithms matched to its detector that allowed Bridger to significantly improve the signal-to-noise ratio and extend the range of its gas measurements. These are described in the full final report. The performance specifications of the transmitter and receiver combination are shown in Figure 5 (right).
Figure 5. Noise equivalent power specrtum of the 2 µm detector. |
|
2.3. Objective 3. Demonstrate Performance
After the transmitter and receiver subsystems were fully tested and integrated, Bridger demonstrated the combined systems critical ability to detect CO2. All of the tasks for this objective were completed, except that Bridger focused on completing one SNR/sensitivity measurement sets instead of the originally planned three.
Figure 6 (left) shows the atmospheric backscattering returns from a 10 minute averaged ambient atmospheric CO2 DIAL measurement performed at Bridger's facility in Bozeman, MT. The figure shows the on-line and off-line average return light as a function of range (calculated from the time of flight of the laser pulses). The figure also shows how Bridger's deconvolution algorithm compensates for the undesired, but unavoidable dip of the return signal below zero for ranges above about 40 m (due to the detector impulse response). Under this contract, the company also developed a comprehensive model that accurately predicts the observed return light levels. Figure 6 (right) shows the resulting CO2 concentration measurement derived from the data, which are in good agreement with nominal current atmospheric CO2 concentration of 390 ppm. Bridger demonstrated a sensitivity of <100 ppm out to 50 m, with the best sensitivity measured near the peak return light from Figure 6 (left). This level of sensitivity is 2 percent of human breath concentration and <1 percent of typical industrial emissions, indicating that this system will be valuable for EPA to remotely test and monitor pollution sites.
Figure 6. Left: The atmospheric bascscattering returns from a 10 minute averaged CO2 DIAL measurement. Right:
The CO2 concentration measurement derived from inverting the data using the DIAL equations.
2.4. Objective 4 and Commercialization Supplement. Design Prototype Device
Bridger's original plan was to perform design work on, but not to construct, a final sensor. However, once Bridger realized the potential for immediate revenue from the stand-alone laser product (without gas sensing capabilities) for laser ablation applications, it requested and received a Commercialization Supplement to develop the laser into a robust and commercially viable product. Bridger overcame serious damage problems and developed a monolithic method of securing its components for excellent robustness and lifetime using the supplementary funds. Bridger is happy to report that it has already sold nine of these lasers and anticipates selling up to ten more in 2013. The revenue Bridger has generated from these laser sales already surpasses the EPA's investment in this project. Bridger has begun and will continue to use revenue generate from these lasers to advance the full gas sensor to commercialization.
Conclusions:
Bridger successfully demonstrated a breadboard prototype CO2 sensor based on differential absorption LIDAR that could perform both integrated column and range resolved concentration measurements. The sensor is capable of detecting atmospheric CO2 to with better than 100 ppm out to 50 meters. Moreover, the company was able to leverage research performed on this contract to develop a commercially available compact and efficient mid-infrared laser source for other applications. Bridger plans to use revenue generated from these laser sales to help fund the commercialization of gas sensors for both carbon dioxide and methane monitoring.
Supplemental Keywords:
small business, SBIR, EPA, carbon dioxide, CO2, Clean Air Act, air pollution, CO2 monitoring, compact sensor, CO2 emissions, industrial emissions, laser, spatial mapping, LIDAR, point-source, carbon sequestration, hand-held sensor, carbon sequestration monitoring, CO2 sensor, remote sensingSBIR Phase I:
Hand-Held Sensor for Remotely Mapping Carbon Dioxide Pollution Sources | Final ReportThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.