Quantitative Optical gas Imaging (QOGI) Device QL100
- Current State of Art Presented by At Seminar Hosted by U.S. EPA Research Triangle Park, North Carolina June 22, 2015 Introduction Optical Gas Imaging (OGI) technology or Infrared (IR) cameras have been widely used for leak detection. However, they cannot quantify the leaks. Providence has been developing a technology that utilizes existing IR cameras to measure mass leak rate (e.g., lb/hr or g/hr) Quantitative OGI or QOGI. We have completed development of a beta version (Model QL100) based on this technology and we are conducting testing here at the EPA facility today with a QA Plan developed by EPA. Although preliminary results for this technology are very promising, this is still a work-in-progress. QL100 Beta Version USB or wireless connection
Infrared (IR) gas detection camera (currently available) USB Tripod to steady the image QL100 an accessorydevice that can quantify and report the mass leak rate (i.e., lb/hr or g/hr) Working Principle IR images of a leak are analyzed for intensity on a pixel-by-pixel basis Each pixel represents a column of hydrocarbon vapor between the camera and the background Pixel contrast intensity is a function of temperature difference between the background and the plume (T) At a given T, the intensity is proportional to the hydrocarbon molecules in the vapor column Leak rate drives both pixel intensity and number of pixels. Inversely, the combination of the two factors determines leak rate. Calibration Equations
LR = a P + b LR = Leak Rate (g/hr or lb/hr) a, b = constants P = aggregated pixel intensity for the gas plume How Does It Work in the Field?
Use IR camera to survey for leaks. When a leak is detected, connect the QL100 device to the camera (USB or wireless). User enters ambient air temperature, wind speed, estimated distance from the plume to the camera, and expected chemical. QL100 does the rest Collects images for about 30 seconds, uses proprietary algorithms to automatically calculate the mass leak rate in g/hr or lb/hr Provides immediate result in the field Example 1 Flange Flange leak Concrete background Distance of 10 feet
Propane release QOGI Method applied Actual rate: 220 g/hr QOGI result: 228 g/hr Error: 4 % Raw video Absorptive plume (black) FLIR GF300 38mm Example 2 Sky Background
Elevated release point Sky background Distance of 20 feet Propane release QOGI Method applied Actual rate: 205 g/hr QOGI result: 187 g/hr Error: -9 % Raw video Emissive plume (white) FLIR GF300 38mm Summary of Recent Test Results
QOGI Accuracy: -17% to 43% across all leak rates and all 80 tests QOGI accuracy very promising vs Method 21 Comparison with Method 21
Screening Value Correlation Eq. ER (lb/hr) Uncertainty:up to 200% Errors up to 200% could be introduced by not correcting for RF. There are other sources of errors as discussed earlier. Combined Error? Uncertainty:-80% to +300% or worse Based on EPA 1995 Protocol, App. C. Errors in Method 21 Screening
Only concentration is directly measured by Method 21 The size of the leak is not considered Different leak rates could have same concentration, and vice versa Same leak rate(500 scc/min propane) A Method 21 test showed that Conc. On the left is 3 times of Conc. on the right. Small leak area (single point) Large leak area (diffused leak) Factors That Could Impact QOGI Results
Temperature difference (T) between the background and the gas plume (~ambient air) IR camera optical lens IR camera focus Distance Weather conditions - wind speed, wind direction, ambient temperature, sunlight, cloud, rain, etc. Chemical composition Background Many of these factors have been accounted for by Providences proprietary QOGI algorithms Detectable vs. Quantifiable
Screen the image for suitable T BEFORE measurement Provide Data Quality Indicator T Screen Red areas have insufficient T for quantification Raw Image Options To Fill The Gap Find a different viewing angle with different background Test at a different time when T is sufficient Enhance the background Enhanced background Quantifiable Detectable Not quantifiable Path Forward EPA QOGI test series QOGI Test 1 today.
More tests to better characterize the vulnerability and resiliency/robustness of this QOGI product Develop Data Quality Indicators (DQI) to describe the expected accuracy and precision of QOGI results for a given measurement