Accurate gas leak detection using tuneable diode lidar systems
When using spectroscopic diode lidar to detect a gas of interest such as methane (CH 4 ), detecting high concentration pathlength (CPL) values of the gas of interest may not actually indicate a plume or leak. In areas such as a refinery, where ambient gas levels may be routinely high, absorption from ambient gas measured over a long pathlength may produce the same CPL value as a much closer gas leak, leading to a false positive. By collecting and using lidar distance information in tandem with the CPL values determined from the lidar system spectroscopic measurements, the statistical properties of the dataset can be evaluated, and ambient concentrations of the gas estimated. A lidar dataset with revised values of CPL that adjusts for these measurements of ambient concentration may be created to provide better reliability for analysis and in for triggering alarms for potentially dangerous gas concentrations and leaks.
1 . A method of creating a lidar image for detecting a gas, the method comprising:
in a lidar system comprising a diode laser driven by a drive current and a detector, changing the drive current of the diode laser to produce laser light over a spectrum of wavelengths, wherein points in time have a relationship to the drive current and the spectrum of wavelengths;
modulating the laser light with a pattern of pulses;
transmitting the modulated laser light at a plurality of angles into a scene;
for each of the plurality of angles, (i) detecting returning laser light scattered from scattering points in the scene over a time interval T, (ii) estimating a corresponding distance d to the scattering points in the scene, and (iii) estimating a corresponding concentration path length (CPL) value for a gas of interest;
storing data corresponding to each angle, of the plurality of angles, along with the corresponding estimated distance and the corresponding estimated CPL value in a first lidar dataset;
calculating, based on the stored data in the first lidar dataset, an ambient gas concentration (C ambient ); and
generating a revised lidar dataset, comprising revised CPL values based on the calculated ambient gas concentration (C ambient );
wherein the estimating of the corresponding distance d to the scattering points in the scene additionally comprises:
for each modulation pulse during the time interval T, generating an intensity histogram based on signals from the detector that correspond to portions of the time interval T, with each intensity histogram having a number of time bins, each time bin
(i) representing a distinct time within the histogram,
(ii) having an index i, and
(iii) having intensity values I i associated with the index i;
combining data for intensity histograms collected during the time interval T into a grouped histogram with a grouped intensity signal I i (τ i ) for each ith time bin;
computing a time t scene for the grouped histogram from time bins corresponding to the detected returning laser light scattered from scattering points in the scene; and
computing the distance d corresponding to the time t scene .
2 . The method of claim 1 , wherein
the spectrum of wavelengths corresponds to an absorption feature for the gas of interest.
3 . The method of claim 1 , wherein
the gas of interest is methane (CH 4 ).
4 . The method of claim 1 , additionally comprising:
providing the pattern of pulses using control electronics;
initiating data collection in an intensity histogram with a signal from the control electronics correlated with a pulse from the pattern of pulses used to modulate the laser light; and
determining a time value to for the intensity histogram that corresponds to the signal from the control electronics; and wherein
the computing of the distance d comprises using
d
=
c
(
t
scene
=
t
o
)
2
,
where c is the speed of light.
5 . The method of claim 1 , additionally comprising:
detecting laser light scattered internally within the lidar system;
determining a time value t 0 for a time bin corresponding to the detected laser light scattered internally within the lidar system; and
determining a time value t scene for N time bins corresponding to detecting laser light scattered from the scene,
wherein the computing of the time t scene comprises using a computation using weighting factors given by:
t
scene
=
∑
i
=
1
N
I
i
(
τ
i
)
τ
i
∑
i
=
1
N
I
(
τ
i
)
where I i (τ i ) is the grouped intensity signal corresponding the ith time bin; and
computing the distance d comprises using
d
=
c
(
t
scene
=
t
o
)
2
,
where c is the speed of light.
6 . The method of claim 5 , wherein the detector detects both:
the laser light scattered internally within the lidar system, and
returning laser light scattered from scattering points in the scene.
7 . The method of claim 1 , additionally comprising
displaying the revised CPL values in the revised lidar dataset in a two-dimensional (x,y) image format, wherein each (x,y) point corresponds to a revised CPL value for a corresponding angle coordinate, and wherein the display uses color coding to represent different magnitudes of revised CPL.
8 . The method of claim 1 , additionally comprising
when one or more revised CPL values in the revised lidar dataset exceeds a predetermined threshold CPL value,
generating a gas leak alarm signal.
9 . A method of creating a lidar image for detecting a gas, the method comprising:
in a lidar system comprising a diode laser driven by a drive current and a detector, changing the drive current of the diode laser to produce laser light over a spectrum of wavelengths, wherein points in time have a relationship to the drive current and the spectrum of wavelengths;
modulating the laser light with a pattern of pulses;
transmitting the modulated laser light at a plurality of angles into a scene;
for each of the plurality of angles, (i) detecting returning laser light scattered from scattering points in the scene over a time interval T, (ii) estimating a corresponding distance d to the scattering points in the scene, and (iii) estimating a corresponding concentration path length (CPL) value for a gas of interest;
storing data corresponding to each angle, of the plurality of angles, along with the corresponding estimated distance and the corresponding estimated CPL value in a first lidar dataset;
calculating, based on the stored data in the first lidar dataset, an ambient gas concentration (C ambient ); and
generating a revised lidar dataset, comprising revised CPL values based on the calculated ambient gas concentration (C ambient );
wherein the calculating of the ambient gas concentration (C ambient ) comprises:
for the estimated CPL values stored in the first lidar dataset,
determining a standard deviation for the estimated CPL values using a noise model;
generating a filtered lidar dataset from the first lidar dataset by removing one or more estimated CPL values, of the estimated CPL values, that exceed a predetermined threshold CPL deviation value;
calculating a functional relationship for CPL and distance d using the estimated distances and estimated CPL values remaining in the filtered lidar dataset; and
determining a value for the ambient concentration (C ambient ) of the gas of interest using the calculated functional relationship.
10 . The method of claim 9 , wherein
the calculating the functional relationship comprises linear regression.
11 . The method of claim 9 , wherein
the predetermined threshold CPL deviation value is at least 3 times the standard deviation for the estimated CPL values for the first lidar dataset.
12 . The method of claim 9 , wherein
the revised CPL values in the revised lidar dataset are calculated from the estimated CPL values in the first lidar dataset by subtracting an amount based on distance d and the determined C ambient from each of the estimated CPL values stored in the first lidar dataset.
13 . The method of claim 9 , wherein
the spectrum of wavelengths corresponds to an absorption feature for the gas of interest, and the gas of interest is methane (CH 4 ).
14 . A lidar system, comprising:
a diode laser, driven by a drive current;
a modulator;
a detector; and
a processor, the processor having associated non-transitory memory programmed with instructions that allow the system to:
change the drive current of the diode laser to produce laser light over a spectrum of wavelengths, wherein points in time have a relationship to the drive current and the spectrum of wavelengths;
direct the modulator to modulate the laser light with a pattern of pulses;
transmit the modulated laser light at a plurality of angles into a scene;
for each of the plurality of angles, (i) detect returning laser light scattered from scattering points in the scene over a time interval T, (ii) estimate a corresponding distance d to the scattering points in the scene, and (iii) estimate a corresponding concentration path length (CPL) value for a gas of interest;
store data corresponding to each angle, of the plurality of angles, along with corresponding estimated distance and the corresponding estimated CPL value in a first lidar dataset;
calculate, based on the stored data in the first lidar dataset, an ambient gas concentration (C ambient ); and
generate a revised lidar dataset, comprising revised CPL values based on the calculated ambient gas concentration (C ambient );
wherein the estimating of the corresponding distance d to the scattering points in the scene additionally comprises:
for each modulation pulse during the time interval T, generating an intensity histogram based on signals from the detector that correspond to portions of the time interval T, with each intensity histogram having a number of time bins, each time bin (i) representing a distinct time within the histogram, (ii) having an index i, and (iii) having intensity values I i associated with the index i;
combining data for intensity histograms collected during the time interval T into a grouped histogram with a grouped intensity signal I i (τ i ) for each ith time bin;
computing a time t scene for the grouped histogram from time bins corresponding to the detected returning laser light scattered from scattering points in the scene; and
computing the distance d corresponding to the time t scene .
15 . The lidar system of claim 14 , additionally comprising
a display configured to present the revised CPL values in the revised CPL dataset in a two-dimensional (x,y) image format, wherein each (x,y) point in the image corresponds to a revised CPL value for a corresponding angle.
16 . The lidar system of claim 14 , wherein the gas of interest is methane (CH 4 ).
17 . The lidar system of claim 14 , additionally comprising control electronics configured to:
provide the pattern of pulses; and
provide a signal to initiate data collection in an intensity histogram, the signal correlated with a pulse from the pattern of pulses used to modulate the laser light.
18 . The lidar system of claim 17 , wherein:
the non-transitory memory is additionally programmed with instructions that allow the system to:
determine a time value t 0 that corresponds to the signal from the control electronics;
compute a time t scene corresponding to the detected returning laser light scattered from scattering points in the scene; and wherein
estimating the corresponding distance d uses
d
=
c
(
t
scene
=
t
o
)
2
,
where c is the speed of light.
19 . The lidar system of claim 14 , wherein:
the non-transient memory is additionally programmed with instructions that allow the system to detect laser light scattered internally within the lidar system, wherein the laser light scattered internally within the lidar system is scattered from a beamsplitter; and
the detector is positioned to detect both:
the laser light scattered internally within the lidar system and
the returning laser light scattered from scattering points in the scene.
20 . The lidar system of claim 14 , additionally comprising:
an alarm mechanism to transmit an alarm when one or more revised CPL values exceed a predetermined threshold.