NEIC ALTERNATE METHOD TO REFERENCE METHOD 9 ALTERNATE METHOD 1 - DETERMINATION OF THE OPACITY OF EMISSIONS FROM STATIONARY SOURCES REMOTELY BY LIDAR v>EPA national enforcement investigations center denver federal center bldg 53. box 25227 denver, co 80225 ------- Environmental Protection Agency Office of Enforcement ALTERNATE METHOD TO REFERENCE METHOD 9 ALTERNATE METHOD 1 - DETERMINATION OF THE OPACITY OF EMISSIONS FROM STATIONARY SOURCES REMOTELY BY LIDAR National Enforcement Investigations Center Denver, Colorado ------- PROPOSED ENVIRONMENTAL PROTECTION AGENCY NEW METHOD FOR STANDARDS OF PERFORMANCE FOR STATIONARY SOURCES ENVIRONMENTAL PROTECTION AGENCY (40 CFR Part 60) Proposed New Alternate Method to Reference Method 9 On December 23, 1971, the Environmental Protection Agency promulgated standards of performance for five categories of stationary sources under Section III of the Clean Air Act, as amended. An appendix to the regulation contained Reference Methods 1 through 9, which detailed requirements for performance testing of stationary sources. Since the promulgation of these reference methods EPA has continued to participate in the development of new, improved and/or supplemental methods that will permit the quantitative monitoring of stationary source emissions remotely (non-contact methods), quickly, easily and with a high degree of accuracy during both day- and night- time hours. A new test method has been developed from a detailed evaluation of scientific remote sensing instrumentation for the determination of station- ary source emissions (plume) opacity. This new method (Alternate Method 1) employs a lidar (lasar radar) which is an optical system installed in a truck-mounted van enclosure. This mobile lidar is specifically designed and fabricated for measuring the opacity of visible emissions from a given stationary source during both day- and night- time ambient lighting conditions. The laser transmitter within the lidar emits a short pulse of light. The lidar receiver collects the laser light backscattered from the atmospheric aerosols before and beyond the visible plume as well as that from the aerosols (particulates) within the plume. The lidar receiver converts the backscattered optical signals to electronic signals or data. From the data (just before and beyond the plume) the plume opacity is subsequently calculated and printed in hard copy format along with the date and time to the nearest second. The nominal data gathering rate of the lidar is an opacity measurement or determination once every 10 seconds. ------- This rate can be maintained in continuous operation for minutes or even hours. The lidar has an inherent data processing capability that calculates the opacity values, subsequently records these opacity values in hard copy form, and also records the original lidar receiver data (signal amplitude vs range) on magnetic tape for future reference or use. This method is intended for use during night-time hours as it is during the day. This is a most impor- tant capability in the measurement of stationary source emissions opacity. By this notice, the Administrator is inviting comments on the proposed new test method. Submittals should, wherever possible, be supported with data and calculations. Comments on the proposed new test method should be submitted, in trip- licate, to the National Enforcement Investigations Center, U.S. Environmental Protection Agency, P. 0. Box 25227, Denver, CO 80225, Attention: Mr. Arthur W. Dybdahl. All comments post-marked no later than 1980, will be considered. Copies of comments received will be available for public inspection during normal business hours at the Public Information Reference Unit (EPA Library), Room 2922, 401 M Street, S.W., Washington, DC. This new test method is proposed under the authority of Section III of the Clean Air Act, as amended (42 U.S.C. 1857c6) Dated: , 1980 Administrator It is proposed to amend Part 60 of Chapter I of Title 40 of the Code of Federal Regulations by adding Alternate Method 1 to Reference Method 9 as follows: ------- ALTERNATE METHOD TO REFERENCE METHOD 9. ALTERNATE METHOD 1 - DETERMINATION OF THE OPACITY OF EMISSIONS FROM STATIONARY SOURCES REMOTELY BY LIDAR Many stationary sources of air pollution discharge particulate emissions into the atmosphere as visible emissions usually as a plume emitted from a stack or other source structure. This proposed method provides the quantita- tive determination of the opacity of an emissions plume remotely by a Mobile Lidar System (laser radar; Lujht Detection and Ranging). The method includes procedures for the calibration of the lidar and procedures to be used in the field for the lidar determination of plume opacity. The lidar is used to measure plume opacity during either day- or nighttime hours since it contains its own optical energy source or transmitter. The operation of the lidar is not dependent upon ambient lighting conditions (light, dark, sunny or cloudy). The most common laser transmitter is the ruby laser which emits red light at 6943 Angstroms. The ruby laser has been chosen for the following reasons: The red light is not absorbed by atmospheric gases including water vapor. The optical attenuation (extinction) of the red light as it passes through particulates in a smoke plume is slightly less than for green or white light. The opacity of a given plume would be slight- ly less for red light than for that measured with green or white 1ight. ------- 2 The ruby laser is very reliable. There is a large amount of technical information regarding the optical properties of the atmosphere as measured and monitored with the ruby laser. The lidar mechanism or technique is applicable to measuring smoke plume opacity at numerous wavelengths of laser radiation. However, the performance evaluation and calibration test/results given in this method apply only to a ruby lidar. These tests were performed using the EPA-NEIC Omega-1 lidar. The lidar has two basic integrated constituents, the laser transmitter and the electro-optical receiver. The laser transmits an extremely short pulse (nearly 5 meters in length) of light toward a visible emissions plume. This light pulse is partially backscattered to the lidar receiver by aero- sols (particulates) from three distinct regions along the lidar's line-of- sight or instantaneous field of view: 1. The atmospheric path before the pulse reaches the plume, 2. The plume itself, 3. The atmospheric path beyond the plume, after the pulse has passed through the plume. The backscattered light signals from regions (1) and (3), corrected for 1/R2 fall-off (optical backscatter signal amplitude decrease as a func- tion of lidar range through the atmosphere), are the optical data used to calculate the optical opacity of the plume under investigation. By the mathematical ratio of the signal from region (3) to that from region (1), the lidar measures the square of the optical transmittance (Tp2) of the plume. The square of the plume transmittance is measured because the lidar pulse effectively passes through the plume twice. The lidar pulse goes out through the plume and subsequently is backscattered by the atmospheric aero- sols beyond the plume [region (3)]. This backscatter signal then returns through the plume to the lidar receiver. ------- 3 The backscatter signals are converted to electronic signals in the lidar receiver. These signals are directed into the data processing intru- 2 mentation which calculates the square of the plume transmittance (T ), the plume transmittance (T ) and then the plume opacity, 0 , 0 = 1 - Tp. The opacity value for each lidar measurement or firing is permanently recorded in a hard-copy format along with its respective date and time data. The original lidar receiver data (optical backscatter signal amplitude vs. lidar 1ine-of-sight range) along with date, time, source identification, etc. are recorded on the system's computer-controlled magnetic tape recorder for future reference or additional calculations. The most important advantages of the lidar for field use are: Its inherent, absolute accuracy in measuring the opacity of a plume, being significantly greater than that obtained with the Reference Method, Its capability of measuring plume opacity during nighttime hours as well as during daylight conditions, which cannot be effectively accomplished with the Reference Method. Its inherent capability of measuring plume opacity with consistent accuracy and nonsubjectivity independent of back- ground light contrast conditions such as between a plume and clear sky, cloudy sky or terrain background, etc. The color contrast between the plume under test and the background sky or terrain has no bearing on the lidar's performance since the only data required is the atmospheric optical backscatter signals from just before the plume and just beyond or behind the plume. If the lidar line-of-sight terminates against either terrain or a cloudy sky, this will not affect the lidar opacity measurements. However, the lidar cannot make accurate opacity measurements while looking directly into the sun or during precipitation con- ditions. ------- 4 While measuring plume opacity of a white-to-gray plume the reference method has a significant negative bias, as documented in the introduction of Method 9, due to the lower contrast between the plume and the background (haze or clouds). Also the opacity error will further be increased as the ambient lighting level decreases toward darkness. Since the measurement of plume opacity with the lidar is independent of plume/background contrast and ambient lighting conditions, the signifi- cant negative bias and negative errors inherently associated with the refer- ence method is not present in the lidar opacity measurements. The lidar mechanism measures the actual plume opacity with greater accuracy than does the reference method. By definition it is usual that the alternative method gives a negative bias (lower value and possibly less accurate) for a given test parameter or variable with respect to the reference method; however, with the lidar mechanism this is not the case. Under less-than-ideal background-to-plume color/luminiscent contrast conditions the reference method cannot be effectively used to verify the data obtained with this method because of the significant negative bias and negative errors. The same holds true with using the lidar to measure plume opacity at night. The reference method cannot be used to verify the data obtained under this method. It is suggested that an industrial facility, etc., use a white-light transmissometer, properly positioned, calibrated, and operated, to verify the opacity values concurrently recorded with the lidar. This is especially suggested during nighttime operations. (Some new source performance stand- ards now require on-stack transmissometers to measure opacity). The calibration tests showed that the lidar (used for these tests) measured opacity with an overall accuracy of 0.2% (standard deviation = 0.6% for 2,880 data values) in the linear channel over the opacity range of 0 to ------- 5 80% (nominal), and from -0.3 to +0.1% (standard deviation = 0.5% for 1,950 data values) in the logarithmic channel from 20 to 80% (nominal). The in- ternal calibration mechanism (optical generator in the 0mega-l Lidar) used to calibrate the lidar in the field was employed for these tests (Reference 5.1). The lidar was subjected to performance evaluation tests with a smoke generator that is used in the certification of federal and state visible emissions observers in accordance with Method 9 requirements. The analyzed opacity data showed that the lidar opacity values ranged from 0% difference to -2%, with respect to the smoke generator transmissometer, for 80% of the reduced data runs. For 93% of the reduced data runs the difference in plume opacity ranged from + 1% to -2%. For about 7% of the reduced data runs the lidar opacity was slightly greater than the transmissometer value by 4% or less. In these latter data the positive error was due to ambient dust, being generated by vehicles operating nearby, present in the near region of the lidar's 1ine-of-sight. However, the data were retained because the standard deviation of the lidar opacity values were less than the 8% limit set forth in the Opacity Data Acceptance/Rejection Criterion (Section 2.4.3 of this Method). The calibration and performance evaluation tests clearly demonstrate that the lidar is an acceptable alternate method. The required correlation (performance evaluation tests) was not carried out with visible emissions observations, due to the inherent negative bias, but with the smoke genera- tor's white-light transmissometer which is routinely used to certify the visible emissions observers under the reference method. The operation of the lidar in the field is performed in accordance with stringent laser safety standards which EPA developed using the refer- ences given in Section 5, as a basis (Section 7 of Reference 5.1). 1. Principle and applicability 1.1 Principle. The opacity of visible emissions from stationary sources (stacks, ducts, etc.) is measured remotely by a mobile lidar (laser radar). ------- 1.2 Applicability. This method is applicable for the remote measurement of the opacity of visible emissions from station- ary sources during nighttime as well as daylight conditions, pursuant to 40 CFR § 60.11(b). It is also applicable for the calibration of the mobile lidar for the measurement of the opacity of emissions. A performance/design specification for a basic lidar system is also incorporated into this method. Procedures. The mobile lidar calibrated in accordance with Para- graph 3 of this method shall use the following procedures for remotely measuring the opacity of stationary source emissions: 2.1 Lidar position. The lidar shall be positioned at a distance from the source-under-test, sufficient to provide an unobstruc ted view of the source emissions. The source-under-test must be at a range greater than the lidar's transmitter/receiver convergence distance along the line-of-sight or field-of-view. The lidar observations shall be carried out in such a way that a good quality, clear air (atmospheric backscatter) re- turn signal, being free of obstructions, is obtained along the lidar's line-of-sight before reaching the plume (near- region) and beyond the plume (far-region). These two signal segments are used to calculate the plume opacity for each lidar measurement. When there is more than one source of emissions in the immediate vicinity of the plume under test, the lidar shall be positioned so that the lidar beam passes through only a single plume, free from any interference of the other plumes for a minium of 50 meters on each side of the plume under test along the line-of-sight. The lidar shall be positioned so that the lidar line-on-sight is nearly per- pendicular to the direction of horizontal drift of the plume, to the extent practicable. This practice will keep the lidar line-of-sight distance through a plume equal to the actual ------- 7 plume thickness, at the point of testing within the plume. The lidar shall be in a position so that direct solar radia- tion is not permitted to enter the optical receiver. When measuring the opacity of emissions from rectangular outlets (e.g., roof monitors, open baghouses, noncircular stacks, etc.) the lidar shall be placed in a position so that the line-of-sight is approximately perpendicular to the longer (major) axis of the outlet as practicable. 2.2 Lidar observations and measurements. Once placed in its proper firing position, the lidar shall be aimed (by aiming telescope) and fired immediately below the outlet of the emissions source in order that the range to this source be measured and recorded, and to assure that the lidar trans- mitter beam diameter at the source range is less than the source (outlet) diameter. The lidar transmitter beam dia- meter must be less than the plume diameter at the particular point or location of test, to conduct the remote opacity measurement. This determination is accomplished in two ways. First the operator shall directly view the reflected lidar pulse through the receiver telescope for complete spatial coincidence and by a lidar opacity calculation which will be 100% (opaque source structure) if the beam diameter is less than the plume diameter. Secondly, the lidar beam diameter shall also be calculated as the mathematical product of two parameters; namely, the inherent beam divergence angle and the lidar range to the stack under test. If the lidar beam diameter is larger than the diameter of the plume under test, then the instrument must be moved closer to the source until the above condition is fulfilled. ------- 8 The lidar is then aimed and fired with the 1ine-of-sight near the outlet height and rotated horizontally in an upwind direction to a position clear of the source structure and the associated plume (if wind conditions are calm, then the lidar may be moved to either side of the plume that is free of obstructions). This lidar data signal is the clear-air return or reference signal. The lidar operator shall inspect this signal on the oscilloscope display to determine if the lidar line-of-sight is free of obstructions, obtain a measure of the integrity and homogeneity of the clear air in front of and behind the source/plume, and to verify that the source/ plume under test is free from interference from other plumes, for a minimum of 50 meters on each side of the plume under test along the line-of-sight, that may be in the immediate area. This clear-air (backscatter) signal shall be recorded on magnetic tape and shall be used in the plume opacity calcu- lations. The frequency of the reference measurements is established in the Temporal Criterion of Section 2.2.3 of this method. Finally, the lidar is aimed at the region of the plume which displays the greatest opacity, where condensed water vapor is not present. The lidar is fired through the plume so that the lidar operator may make any final aiming adjustments and choose or verify the proper locations of the backscatter signal segments used in the opacity measurement. When that is completed, the lidar is placed in operation with the nominal pulse or firing rate of 6 pulses-per-minute (pulse/10 seconds as safety permits). The lidar operator shall observe the oscilloscope display to monitor lidar performance and the quality of the plume data signals. (The display provides a near real-time means of determining the quality and integrity of the near- and far-region atmospheric backscatter signals ------- 9 used in calculating plume opacity). These plume data signals are recorded on magnetic tape as a permanent record. The plume data signals recorded from lidar start to stop is called a data run. Short term stops of the lidar to record addition- al reference measurements do not constitute the end of a data run if lidar operations for a given plume are resumed within 60 seconds after the reference measurement has been recorded. The temporal length of an individual data run may extend from 1 or 2 minutes, such as for intermittent sources, to over an hour or even longer depending usually upon the cha- racteristics and variability of the source emissions. The lidar data rate is nominally maintained at one opacity mea- surement every 10 seconds throughout a given delta run. The lidar will be used to measure the opacity of hydrated or so-called steam plumes. (To the extent practicable the lidar operators should have technical information with them regard- ing the respective process and the control equipment for each stationary source to be listed.) As listed in the refer- ence method there are two types, i.e., attached and detached steam plumes. 2.2.1 Attached Steam Plumes: When condensed water vapor is present within a plume-under-test as it emerges from the emission outlet, the opacity measurements shall be made with the lidar at a point within the residual plume where the condensed water vapor is no longer visible. During daylight hours the lidar operator can usually locate the most dense portion of the residual plume visually. The operator shall then aim the lidar transmitter/receiver into ------- 10 that portion or region of the plume. During either day- or nighttime operations the lidar is used to locate the most- dense region of the residual plume, i.e., the region of high- est opacity. (A high intensity spotlight, night vision scope, or low light level TV, etc., can be used to aid the lidar operator in aiming the transmitter/receiver at night). The lidar operator scans the transmitter/receiver, with the lidar measuring opacity, along the longitudinal axis or center line of the plume from the emissions outlet to a point just beyond the steam plume. The steam plume will have nearly 100% opacity while the residual plume opacity is most probably much lower. If the residual plume also has a 95 to 100% opacity then the lidar operator shall also observe color differences as an added assurance that the lidar is aimed completely within the residual plume. Plume reflectivity can also be used to accomplish this same task, especially at night. The steam plume is white and highly reflective while the residual plume will be lower in reflectivity. Once the residual region of the plume is located (along its center line) the lidar transmitter/receiver shall then be scanned approximately perpendicular to this axis in order to locate the region of highest opacity. Plume opacity shall then be measured at this location within the plume. Adjust- ments are made to this location of the lidar 1ine-of-sight within the residual plume as deemed necessary by the lidar operators to correct for changes in wind direction, etc. The distance from the stack to the position within the plume where these opacity measurements are collected shall be ob- tained by a calculation using the lidar range to the stack, the lidar range to the plume monitoring position, and the ------- 11 azimuth/elevation angles between the stack and the plume monitoring position or point. The geometry for this calcu- lation, along with the respective variables, is defined in Figure AM1-1. This distance is calculated by using Equation AM1-1. 2 2 R.. = [R +R -2R R (Cos p Cos p Cos i|> (AM1-1) SiJ> L p s ps Kp + Sin p Sin p ]2 p s The variables Rs, Rp, Ps, and ip are measured directly with the lidar. The three angles are measured on the trans- mitter/receiver mount by using an inclinometer and a protrac- tor. The two range values are calculated from the respec- tive backscatter signals. R^^ shall be calculated and record- ed for each postion of the lidar line-of-sight while the tests are being conducted. 2.2.2 Detached Steam Plumes: When the water vapor in a hydrated plume condenses and becomes visible at a finite distance from the stack or source emission outlet, the opacity of the emissions shall be measured in a region of the smoke plume just above the emissions outlet prior to the condensation of the water vapor. The condensation of the water vapor in the source emissions forms the steam plume which appears white, and is usually about 100% opacity. During daylight hours the lidar operator can visually deter- mine if the steam plume is detached from the stack outlet. At night a high intensity spotlight, a night vision scope, low light level TV, etc. can be used as an aid in determining if the steam plume is detached. The lidar is also used to determine if the steam plume is detached from the emissions outlet by repeatedly measuring plume opacity, from the outlet ------- 12 2 (plume measurement position) Lidar Position Projection of ' onto the xy-plane Projection of Pp onto the xy-plane The mathematical variables or functions are defined as follows: = the range from the emissions point, P , to the plume monitoring point or position, P . R$ = the range from the lidar to the source or stack outlet. = the elevation angle of the lidar 1me-of-sight above the horizontal plane, to the stack outlet. Rp = the range from the lidar to the plume monitoring point. B = the elevation angle of the lidar line-of-sight above the horizontal plane, to the plume monitoring point. i|> = the azimuthal angle, in the horizontal plane, of the lidar line-of-sight measured from the y-axis which contains the lidar and stack outlet. P (R , 0, 6 ) = Coordinates of the stack outlet in the spherical coordinate system. P_(R , B ) = Coordinates of the plume monitoring point in the spherical p P coordinate system. Figure AM1-1. Geometry of the Residual Plume Measurement with Lidar ------- 13 to the steam plume along its longitudinal axis or center line, and/or observing plume reflectance. Once the determination of a detached steam plume has been confirmed, the lidar is then aimed into the region of the plume between the outlet and the formation of the steam plume, about one half a stack diameter above the outlet. The lidar transmitter/receiver is then scanned across the plume to locate the region of greatest plume opacity. Plume opacity shall subsequently be measured at this location. Adjustments are made to the loca- tion of the lidar's 1ine-of-sight within the plume as deemed necessary by the lidar operators to correct for changes in wind direction, air temperature changes, etc. The location of the lidar's 1ine-of-sight within the plume is recorded for each position while the tests are being conducted. In the measurement of plume opacity from a sulfuric acid manufacturing facility, the lidar 1ine-of-sight shall be positioned within the most-dense part of the plume which will not necessarily be at the emissions outlet. The charac- teristic sulfurous gas absorbs plume moisture forming sulfuric acid aerosols in the submicron size range. High values for the opacity can occur. Often the aerosol plume does not become visible for a few stack diameters away from the emis- sions outlet. This does not constitute a detached steam plume and shall not be treated as such. 2.2.3 Temporal Criterion for Reference Measurements. This criter- ion describes how often and under what conditions additional reference (clear air) measurements are to be obtained. A reference measurement is obtained with the lidar and record- ed on magnetic tape within a 60-second time period prior to any given plume opacity data run. Another reference measure- ment is obtained within 60 seconds after the completion of ------- 14 the same data run. This is standard operating procedure irrespective of the variability of the local atmospheric conditions along the lidar's 1ine-of-sight. The reference measurement shall be obtained by directing the lidar's 1ine-of-sight near the emission's outlet in height or elevation, and rotated horizontally in an upwind direction to a position clear or free of the source structure and the associated plume. If wind conditions are calm, then the lidar 1ine-of-sight may be moved to either side of the plume that is free of obstructions. The need for an additional reference measurement(s) is a function of local atmospheric kinetics which is usually deter- mined through the judgement of the lidar operators as they observe localized meteorological conditions and the character- istics of the lidar backscatter return signals. An additional reference measurement is usually obtained, which occurs during a data run, if the lidar operator (safety observer in the laser room) observes a change in wind direc- tion or plume drift of 30° or more from the direction that was prevalent when the last reference measurement was made. If the lidar operator, stationed in the computer room, ob- serves a noticeable change in the amplitude variations in either the near- or far-region backscatter signal segments (not due to a common change in plume opacity) that remains present for three plume data records (about 30 seconds), then the data run shall be interrupted and another reference measurement shall be recorded. (The location on tape, time, and the proper identity of each reference measurement shall be recorded on magnetic tape). Then the data run shall be immediately resumed and continued through completion. This ------- 15 process of obtaining additional reference measurements may be iterated as many times as required. If the ambient (clear air) conditions along the lidar's 1ine-of-sight are continual- ly changing significantly, then reference measurements and plume data measurements are usually recorded alternately. During the subsequent analysis of the lidar data, the refer- ence and data measurement signals shall be analyzed in the same sequence that they were recorded in the field. Section 2.1 of this method requires that the lidar shall be positioned so that the lidar 1ine-of-sight is nearly perpen- dicular to the direction of the horizontal drift of the plume, to the extent practicable. If, during the course of a data run, the direction of drift of the plume changes by 30° or more, as observed from the lidar positions, and if the lidar 1ine-of-sight is more than about 3 source diameters from the sources outlet while recording plume opacity data, then the operation shall be momentarily interrupted. The parameters Rx, R2 and a of the Elevation/Azimuth Angle Cor- rection Criteria [Section 2.4.4 of this Method] shall be measured and the respective backscatter signals recorded on magnetic tape. Rx and R2 shall be obtained from the plume backscatter signals, measured at the center of each plume return spike [Figure AMl-4(b) of Section 2.4.1]. Usually the need to measure these parameters is conincident with the need for another reference measurement. 2.3 Field Records. The lidar operator shall assign a control number in the lidar logbook for each independent source under investigation even if more than one source is located within the same industrial facility [Figure AM1-2]. This assigned ------- Ill) AH I0(. HUNT II01 M UK Kit TAIULATION (rnni ) lAssijn i CONTROL NUMBER to each individual source undar test) LIDAR LOC CONTROL NUMBER TABULATION Log RooL Number- (Assign a CONTROL NUMBER to each individual source under test] CONTROL NUMBER CONTROL NUMBER DATE ASSI6NED PROJECT CITY, STATE DATE ASSIGNED PROJECT CITY. STATE continued on next page Next Log Book Numtm- Figure AM1-2 Lidar Log Control Number Tabulation en ------- 17 control number shall also be recorded on magnetic data tape for each measurement. (All respective lidar data and reports shall have the control number recorded therein.) The name of the facility, type of facility, emission source type and official designation (stack, open baghouse, etc; stack 001, etc.) and the geographic location of the lidar with respect to the source under-investigation shall be recorded in the lidar logbook [Figure AM1-3]. The date of the test and the time period that a given source was monitored shall also be recorded in the logbook. The 1idar-measured vector (emissions source range and angle referenced to magnetic north) and plume characteristics also are recorded in the logbook. The wind speed, wind direction, air temperature, relative humidity, barometric pressure, visibility (measured at the lidar's position) and cloud cover are recorded in the logbook at the beginning and end of each time period for a given source. A small sketch shall be recorded in the log book depicting the location within the plume of laser beam incidence. The parameters 0 , p and t|i, defined in Figure AM1-1, shall s p be measured at the lidar transmitter/receiver mount and recorded in the log book for each source and for each measure- ment position or location within a given plume under-test. Also the parameters a, Rx, and R2, defined in Section 2.4.4 of this method for the Elevation/ Azimuth Angle Correction Criterion, shall be recorded in the log book for each source, each measurement position or location within the plume and for each instance when the plume drift angle is measured. In addition to the assigned control number, the date, time (to the nearest second) of each measurement, the source ident- ification, the video channel used (linear/ logarithmic channel, for example), the digitizer's sample interval and the location or address of the clear-air reference signal on the magnetic ------- Facility aaaa tad UcatUa IIDAH LOC OF OPERATIONS Control number. 0 VIH. A - LIDAR OPERATOR S NOTES | lac I u d • politico ol luir bun within plum-- attached plgao, etc ) At lit field illt • • II fto« •• (••«¦! !'¦•) locailoe of LIDAR Ok«ctian to source Range to souict km later inclination [~ angle is up horizontal is 0°| Soorco IfM o«d official datlgMolloa Hobo (horoitorlltlci (color, ibopo, ilooa proioat, alf ) Wlod ipood basin hm/hr and km/hr WI¦ d diroitloi begin and Air loaptrotoro t>sgin °C end ° C lolotlio hialdlty begin % end % laioaitar begin end Vlilblllir begin km end km Clootf io»«i begin end Dot* nurdi aodo It Held (topoi, prlotooti, photo I, ole ) MAGNETIC TAPES tapes luckn lilts t ID * t fUNCTIOH >HIM(ATI0H Source optical generator I ) scream I I Data of lait calibration This tait recorded on tap*# tracks 1 2 3 4 5 1 7 8 Calibrated opacity Calculated opacity Recorded on III* OPftATOft'S StGMATUtt DATI OPIIATOl'S JISHATUI1 DAT! WITMSS SIGNATUKI DAII WITNISS SIONATUII DATI: I—' Figure AM1-3 Lidar Log Of Operations 00 ------- 19 data tape shall be recorded on the magnetic data tape for each lidar measurement. If a detached or attached steam plume is present at the emis- sions source under-test, this fact shall be recorded in the lidar logbook along with the quantitative data depicting where in the plume the lidar measurements were conducted [Equation AM1-1]. Each magnetic tape used to record field and calibration data signals shall be assigned a number and that number recorded in the lidar log book. 2.4 Lidar Data Reduction and Records. The lidar data shall be analyzed by computer at the time of each lidar measurement or having been recorded on magnetic tape, shall be analyzed/ verified at a later time. The opacity value and the associ- ated statistics of the emissions from the respective source under-test shall be computed and recorded along with the time and magnetic data tape address for each measurement. 2.4.1 Opacity Calculation and Data Analysis. Referring to the lidar signal amplitude vs. lidar range traces in Figure AM1-4, the opacity value (Op) for each lidar measurement is calculated using Equation AM1-2. 0p = 100% (AM1-2) where: I = near-region signal amplitude, plume data signal, 1/R2 corrected, ------- 20 Convergence Point \ (flat) (negative slnn^ (exaggerated) ZeroSmna^ Level Lidar Range Or Time (a) Clear-air reference measurement made near the plume in order to account for the prevailing non-ideal atmospheric conditions. The signal is 1/R2 corrected Plume Spike (Near Region) (Far Region) Signal Plume Lidar Range Or Time (b) Lidar return (plume data) signal showing the effect of plume attenuation on the backscatter signal in the far region The signal is 1/R2 corrected. 1 Nanosecond-10"® Seconds. Range=Speed of Light Time/2 ()=Pick interval 100 nanoseconds in length Figure AM1-4 Traces Of Lidar Backscatter Signals ------- 21 1^ = far-region signal amplitude, plume data signal, 1/R2 corrected, Rn = near-region signal amplitude, clear-air reference signal, 1/R2 corrected, and R.p = far-region signal amplitude, clear-air refer- ence signal, 1/R2 corrected. The lidar backscatter signal traces shown in Figure AM1-4 are corrected or mathematically compensated for the 1/R2 signal amplitude decrease, where R is the lidar range func- tion. This decrease is inherent to the laser energy signal backscattered from particles and aerosols in the atmosphere. This correction must be performed on each backscatter signal before plume opacity can be calculated. Since in lidar terms, range and time differ only by a constant, one-half the speed of light, the 1/R2 correction for the amplitude decrease may be done in the time domain ( j^2 * R2 = 1, ^2 ' t2 = D- R = c t/2, where R is lidar range depicted in Figure AM1-1, c is the speed of light being 2.9979 • 108 meters/ second and t is lidar time measured from the time the laser was fired, t . 0 The 1/R2 correction mechanism for a backscatter signal is depicted in Figure AM1-5. This backscatter signal having been converted to an electronic signal through the optical detector or photomultiplier tube, was converted from an ana- log electronic signal to a digital electronic signal by a high speed digitizer, necessary for computer processing. This digital signal is comprised of many short segments or time intervals. (On the oscilloscope display the digital signal would appear to the operator identical to the analog ------- Lidar Range Or Time Antp < A2 due to atmospheric extinction R=ct/2, R2=c2t2/4, (1/R2) R2 = 1 £il2 = (1/t2) t2=1 c2t2 4 4 2 Figure AM1-5 1/R Correction Mechanism ------- 23 signal if it were displayed.) The length or size of these time intervals is selected in the digitizer by the operator. The standard size for the time (sample) interval in this application is usually 10 nanoseconds. Each time interval in the digital signal, beyond or later in time than t , is subjected to the 1/R2 correction. The signal amplitude, An> of the nth time interval is multiplied by the square of the time, elapsed from t , defining that interval. In Figure AM1-5, the uncorrected signal amplitude, An is multiplied by the square of the time of the nth interval, t , yielding Antn2, the corrected amplitude. This process is carried out for each time interval in the backscatter signal producing a 1/R2 corrected signal. In Equation AM1-2, I , I -, R and R* are each chosen by the n f n f lidar operator or data analyst. I and 1^ are 100 nanosec in length [Figure AM1-4]. The value of I must be, in time, within the interval chosen for R , and I- must be within n f that for Rf. The lengths of Rn and Rf shall be 100 nanosec minimum. I and R are the plume data and reference pick intervals n n K K for the near region of the lidar backscatter signal, and 1^ and are the respective pick intervals for the far region [Figure AM1-4]. The criterion for the selection of the pick intervals is best described by example. Figure AM1-6 shows 3 actual lidar backscatter return signals which were plotted by com- puter from a field data tape. Figure AMl-6(a) is the 1/R2- corrected clear air reference backscatter signal recorded for use in calculating the plume opacity from the plume signal [Figure AMl-6(b)]. These signals contain slight ------- 24 Convergence Point / (Near Region) (Far Region) 12 10 38 (a) Reference Signal, 1/R Corrected Plume Spike 12 25.59 (b) Plume Data Signal, 1/R Corrected Plume Spike -Atmospheric Backscatter Noise Area 1 (c) Plume Data Signal, 1/R Corrected Area 2 NOTES: (l) Minimum distance from convergence point to the plume spike is 50 meters. (2) All pick intervals are 100 nanoseconds wide. (a) Clear Air Reference Video Signal, l/R2-Corrected, showing slight atmos- pheric noise. This reference signal is for (b). R^, R^ are chosen as indicated coincident with In, 1^. (b) Lidar-return Video Signal, l/R2-Corrected, showing slight atmospheric noise, plume spike and the decrease in atmospheric backscatter signal level in the far region due to the opacity of the plume encountered. I , L are chosen as indicated, n f (c) Lidar-return Video Signal, l/R2-Corrected , showing significant atmos- pheric noise in the near region, plume spike, minimal noise in the far region and the decrease in far region signal level due to the opacity of the plume encountered. In> 1^ are chosen as indicated. Figure AMI - 6. Computer Plots of Lidar A-Scope Backscatter Signals ------- 25 atmospheric backscatter noise as observed in the ripple or variations in amplitude to the right of the point o'>: conver- gence. The near-region pick interval, In, is chosen as close to the plume as practicable with the signal quality in the chosen interval being of minimum overall amplitude and minimal amplitude variation. The reference signal pick interval, R , must be chosen for the same time interval as n I as depicted in Figure AMl-6(a,b). The far-region pick interval, If( is also chosen as close to the plume's far side as practicable. The quality of the backscatter signal in this chosen interval is of minimum overall amplitude and minimal amplitude variation. The far- region reference signal interval, R^, must be chosen over the same interval as 1^ [Figure AMl-6(a,b)]. Figure AMl-6(c) is a computer plot of a lidar backscatter return signal showing significant levels of atmospheric backscatter noise in the near-region which was due to fugi- tive dust blowing in. front of the lidar. In this case there are only two areas in the near region where the pick interval, 1^, can be selected, i.e., areas 1 and 2 as shown. The average signal amplitude over the 100-nanosecond time inter- val in each of these two areas is the same. However, in applying the above criterion, Area 2 is the best interval to be used for the plume opacity calculation (the respective reference signal is not shown). The far-region pick inter- val, 1^, is chosen as shown in Figure AMl-6(c) according to the criterion. Any desired pick interval, such as areas 1 and 2 in Figure AMl-6(c), that is not 100 nanoseconds wide shall not be used in the opacity calculation. If no such interval exists in the near-region or the far-region then the plume backscatter signal shall not be used for the opac- ity calculation. Many additional field-oriented examples of pick interval selection are available in Reference 5.1. ------- 26 Once the pick intervals have been chosen, the respective amplitudes or values of each is calculated. The amplitude of I shall be calculated by the average of the individual amplitudes of the 10 equal time segments or intervals (usual- ly 10 nanoseconds) that comprise the respective pick inter- val (100 nanoseconds in length) according to Equation AM1-3. 1 m ]n = £ 1 (AM1-3) n m . ^ i where: I.j = the amplitude of the ith segment or interval, I = sum of the individual values, m = number of segments in the pick interval (m=10), I = average amplitude of the near-region pick interval. The standard deviation, ST , of this set of 10 individual In amplitude segments (near region) shall be calculated accord- ing to Equation AM1-4. 'in m I i=l (I. - V2 (m-1) (AM1-4) The values of I,, R , R • ST SD , SD„ shall be calculated T n T it Kn KT using this same procedure [Equations AM1-3 and AM1-4]. The plume opacity, 0 , is calculated according to Equation P AM1-2. Opacity in percent is calculated by multiplying the square root of the expression in the brackets of Equation AMI-2 by 100. The standard deviation, SQ, of each opacity value, Op, shall be calculated. It is obtained by a multi-variable function which is given in terms of the standard deviation o1 the ------- 27 individual variables. Given Equation AM1-2 for opacity and the standard deviations previously calculated, the standard deviation of the opacity value is calculated according to Equation AM1-5. where: SQ = standard deviation of the opacity value, Op. 90p/aRn = partial derivative of the opacity function [Equation AM1-2] with respect to the clear-air ref- erence signal variable in the near-region [Figure AM1-6]. Spn = standard deviation of the pick-interval segments for the clear-air reference signal in the near-region. 90p/9Rf = partial derivative of the opacity function with respect to the clear-air reference signal vari- able in the far-region. = standard derivation of the pick-interval segments for the clear-air reference signal in the far-region. 0Op/3In = partial derivative of the opacity function with respect to the plume backscatter signal variable in the near-region. ST = standard deviation of the pick-interval segments for the plume backscatter signal in the near-region. 30p/3I^ = partial derivative of the opacity function with respect to the plume backscatter signal variable in the far-region. Sj^ = standard deviation of the pick-interval segments for the plume backscatter signal in the far-region. The calculated values of 0 , S , I , R , I,, R,, SD , p' o' n' n' f f Rn Snj-, ST and S,c shall be recorded as a permanent record Rf' In If ^ along with the time that the lidar recorded the plume backscatter signal. (AM1-5) ------- 28 2.4.2 Reduction Mechanism for Opacity Data. As given in Section 2.2. the temporal length of an individual data run may extend from 1 or 2 minutes, such as for intermittent sources, to over an hour or even longer depending usually upon the charac- teristics and variability of the source emissions. The lidar data rate is nominally set at one opacity measurement every 10 seconds throughout a given data run. The mechanism by which the set of individual opacity values, Op, comprising a given data run are reduced, is a function ¦v of the air quality regulation to be enforced. When a given state air pollution control regulation specifies a maximum permitted opacity value over a fixed time period (example: Plume opacity shall not exceed 50% for a continuous period of no more than 5 minutes in any 60 consecutive minutes), then that time period or interval shall be used in the reduc- tion of the opacity data. If the respective regulation spec- ifies an opacity limit for an I-minute interval and the data run were I minutes in length, then all the opacity values, measured on the 10 second repetitive cycle or data rate, calculated for this interval shall be averaged yielding an average opacity, Op, for this interval. If the average opac- ity is greater than that permitted by the regulation, then the source is in violation. The average plume opacity, Op, for the I-minute time interval shall be calculated as the average of the consecutive (in time) individual 1idar-measured opacity values, Op, by using Equation AM1-6. (The I-minute time interval is called the "averaging interval"). 1 1 0 = - Z 0 . p l k=i ek (AMl-6) ------- 29 where: 0 , = the kth opacity value in the (I-minute) averaging " i nterval, 1 = the sum of the individual opacity values, 1 = number of individual opacity values contained in the averaging interval, Op = average opacity over the averaging interval. If the respective regulation specifies an opacity limit for an I-minute interval and the data run were J-minute;> in length (J > I), then a running average or progressive average shall be used to reduce the lidar opacity values for a given data run. The mechanism for the running average is shown in Figure AM1-7. The I-minute interval shall be maintained constant in length (temporal) being moved along the entire length of the J-minute data run. Once the opacity values, from 1 to i [Figure AMl-7(a)] have been averaged for the first I-minute time interval by Eq (AM1-6), the running average is performed by successively subtracting the mth value and addinij the n + 1 value [Figure AMl-7(d)] and calculating the averaqe for those i opacity values again, then subtract the m + 1 value and add the n + 2 value and perform the calculation again, etc. The running average is a computational tool which locates the I-minute interval within J that has the highest average opacity. This applies directly to the example given above, i.e., the 5-minute period (1=5) in any 60 consecutive minute period (J=60). The number of values averaged in this manner will not always be equal to a constant i, but the time interval I shall be the same throughout J. A few of the i values may possibly be rejected due to the Opacity Data Acceptance/ Rejection Criterion presented in Section 2.4.3 of this method. ------- 1 2 3 4 ... 1 (a) First average opacity, Op, calculated for the i opacity values 1 2 3 4 . . . i i+1 j (b) Second average opacity calculated, first opacity value subtracted and the (i+1) value added. 12 3 4 —hH r- . l i+1 i+2 (c) Third average opacity calculated, second opacity value subtracted and the (i+2) value added. K- i i i i- 1 2 3 ^ H h + H 1—\- i i+1 i+2 m-1 m m+1 n-1 n n+1 n+2 (d) The mth average opacity calculated. b—i >4 i—i—i—i 1—i—i 1 1—i—i 1 2 3 14 ... i i+1 i+2 j-i ... j-2 j-1 j (e) The last average opacity calculated over the time interval I. *1 is the averaging interval established by State/Local Regulation. Figure AM1-7. Pictoral Diagram of the Running Average. ------- 31 When the applicable control regulation specifies a maximum opacity value as a function of time, then the lidar opacity values, measured on the nominal 10-second data rate, are reduced accordingly by computer. The time intervals over which the opacity values exceed the maximum given in the control regulation shall be summed together within the speci- fied consecutive or overall time period. If the summed time period exceeds the allowable time period the source is in violation. An example of this is the following: Suppose the state regulation states that short-term occurrences shall not exceed 50% opacity from a period aggregating no more than 5 minutes in any 60 consecutive minutes and/or no more than 20 minutes in any 24-hour period. The time intervals over which the plume opacity exceeded 50%, are summod together. If the sum of the intervals exceeds 5 minutes in any 60 con- secutive minutes then the source is in violation. The same holds true if the sum of the individual time intervals exceeds 20-minutes in any 24-hour period. If there is no applicable state air pollution control reg- ulation for the lidar data to be reduced, then the 6-minute time interval of Reference Method 9 shall be used. The running average technique described above, shall be used to calculate the 6-minute interval which has the highest average opacity within a given data run. Referring to Figure AM1-7 I is equal to 6 minutes and J is 6 minutes or longer. The opacity of intermittent visible emission and cyclic proces- ses is measured over a period of time considered adequate to determine compliance/noncompliance with the applicable regu- lation. A cyclic process is defined in Figure AM1-S. 0. = 0% Figure AM1-8 Cyclic Process ------- 32 If the regulation, such as a state or city regulation in an approved state implementation plan (SIP), specifies an opacity limit as a function of time, the 1idar-measured opacity values shall be added together in accordance with the requirements of the regulation. If there is no applicable state or local regulation then the 6-minute interval will be used as described above. If the time period of a given cycle is less than 6 minutes, then the opacity values for this period shall be added to sufficient number of zeros to obtain the 6-minute period. The average opacity is computed from the opacity values and the added zeros. For example, if a particular cycle was 4 minutes in length there would be 24 opacity values (4 minutes x 6 opacity values/minute). Then 12 zeros would have to be added to bring the total to the 36 required values (6 minutes x 6 opacity values/ minute). In support of 40 CFR Part 51 with opacity limits as a function of time, this method shall be employed by summing the respec- tive opacity measurement time intervals for the required period of time. 2.4.3 Opacity Data Acceptance/Rejection Criterion: The plume opacity, Op, is calculated from the lidar backscatter signal data using Equation AM1-2. The standard deviation, SQ, of each respective opacity value is calculated using Equation AM1-5. SQ is an indicator of the quality or integrity of the optical backscatter signal segments from the near-region and the far-region of the lidar 1ine-of-sight [Figure AM1-4], and is termed an atmospheric noise indicator. ------- 33 In the course of reducing large amounts of 1idar-measured opacity data, it was empirically or fundamentally determined that if SQ is greater than 8% (calculated with the plume opacity, 0 , for the selected near-region and far-region P pick intervals), the lidar backscatter signal is not reliable (too noisy) for an accurate opacity measurement. In this case the respective opacity value shall be discarded. For a given data run, if the average of the respective indi- vidual standard deviation values, SQ, of a set of opacity values in an averaging interval, I, is greater than 8% (also based on 100% opacity full scale) then the average opacity, 0p, for that interval shall be rejected and discarded from whole data run. This average is calculated using Equation AM1-7. 1 S = - Z S , (AM1-7) 0 1 k=l 0k where: S . = the kth standard deviation value of the data set I, ok ' I = the sum of the individual standard deviations, 1 = the number of individual standard deviation values in a given data set, SQ = the average standard deviation for a given data set. 2.4.4. Elevation/Azimuth Angle Correction Criterion. To ensure true plume opacity for a given plume under-test, the effect of the elevation angle (angle of inclination of the lidar transmitter/receiver) of the lidar firing through a vertical plume shall be taken into consideration in the opacity calcu- lation. The elevation angle is measured with respect to the horizontal reference line. As shown in Figure AM1-9, the optical plume opacity is typically measured with the lidar ------- Stack's Vertical Axis Vertical Smoke Plume Lidar Elevation or Inclination Angle Horizontal Plane Effective Plume Thickness Actual Plume Thickness LCosB Opacity measured along path L Opacity value corrected to the actual plume thickness, P Lidar Line-of-Sight Referenced to Level Ground (Horizontal Plane) Smoke Stack Figure AMI - 9. Elevation Angle Compensation for Vertical Plumes. CO ------- 35 along the inclined path L. The opacity value ultimately required is along path P, the horizontal thickness of the plume. An individual 1idar-measured plume opacity value, Op, shall be corrected for elevation angle if the lidar transmitter/ receiver elevation or inclination angle, pp, is greater than or equal to the value calculated in Equation AM1-8. (ip>Cos-1[l-M].Opin%, (AMI- If Pp is greater than or equal the expression on the right side of Equation AM1-8 then the opacity value, Op measured along the lidar path L shall be mathematically modified or corrected to obtain the opacity value, 0pC, for the actual plume (horizontal) path or thickness P, by using Equation AM1-9. 0 = 0 Cos p (AM1-9) pc p Hp This correction keeps the maximum difference of (0 - 0 ) p pc to approximately 1% (full scale). A given 0pC shall be used in place of its respective Op in the Reduction Mechanism [Section 2.4.2]. When measuring the opacity in the residual region of an attached stream plume [Section 2.2.1 of this Method], the lidar shall be positioned in relation to the stack so that the lidar 1ine-of-sight is nearly perpendicular to the direction of the horizontal drift of the plume, to the ex- tent practiable [Section 2.1 of the Method]. This proced- ure will essentially keep the lidar 1ine-of-sight distance through the plume equal to the actual plume thickness at the point of opacity measurement. However, if the direc- tion of drift of the plume should change so that the lidar ------- 36 1ine-of-sight does not pass through the plume nearly perpen- dicular, then an azimuthal angle correction shall be made to the calculated opacity values obtained under this condition. The geometry of this correction is defined in Figure AM1-10. This correction shall also apply to the other plume types if the opacity measurement is made more than 3 source diameters away from the source outlet. The drift angle, e, is obtained from Equation AM1-10. e = Cos -l R2 . R2 1 r 2 RjR A R2 2 (AMI-10) where: Rx = the lidar range to the position within the resid- ual plume where opacity measurements are being performed, Position 1 in Figure AMl-lO(a). a = Azimuthal angle through which the lidar transmitter/ receiver is turned in order to measure the drift angle; a > 5° as measured on the lidar transmitter/receiver mount. R2 = the lidar range to the position within the plume selected in order to measure the drift angle, Postion 2 in Figure AM1- 10(a). R^ = the distance along the center line of the plume, in the direction of drift, from Position 1 to Postion 2 [Figure AMl-lO(a)]. Ri and R2 are measured directly from the respective plume backscatter signals at the center of the plume spike. The angle a is measured at the pedestal of the lidar transmitter/ receiver. ------- Position of the Lidar within residual plume measurement [Position 1i ne-of-sight for opacity !]• Plume Direction of Drift perpendicular to the Lidar 1ine-of-sight at initial setup of the Lidar (along plume center line); e = tt/2. Plume Direction of Drift later in time, not perpendicular to the Lidar 1ine-of-sight (along plume center line). Position within the plume selected in order to measure the plume drift angle [Position 2]. Lidar Position (a) Plume Dr Lidar Line-of-Sight, Position 1 (b) Figure AMI-10. Correction in Opacity for Drift Residual Region of an Attached Steam Plume. ------- 38 If e > 100° or e < 80° for Op in the range from 50% to 100%, if e > 105° or £ < 75° for Op in the range from 20% to 40%, and if e > 120° or e < 60° for Op in the range from 1% to 20%, then the azimuthal correction shall be performed on the lidar measued plume opacity value 0 using Equation AM1-11. P 0 = 0 Cos ( ~ - e) = 0 Sin £ (AM1-11) pc p 2 p where: Op = the opacity value measured along the lidar path L', which is the thickness of the plume along the lidar line-of- sight through the plume, Position 1 in Figure AMl-10(b). 0pC = the actual plume opacity along the corrected path P' [Figure AMl-10(b)]. A given 0 shall be used in place its pc respective 0 in the Reduction Mechanism [Section 2.4.2]. P There may be testing situations where both the azimuth and the elevation corrections shall be performed. In this case the elevation angle correction is made first and then the azimuth angle correction is carried out on the opacity value already corrected for elevation. 2.4.5 Lidar Data Analysis Record. While the lidar data analysis and reduction are being conducted permanent records shall be initiated and maintained. In these records, which may be a laboratory log book or the paper output from a computer printer, the measured or calculated values for I , Sjn; 1^, SIf; Rn' SRn' Rf' SRf' Rs' Pp' ^ R6i|i: Rl* Rz' a' e; °p' S0, 0pC, along with the respective units (meters, nanoseconds, etc), shall be recorded for each final opacity calculation. It shall be clear, from these records, what data processing operations were used to calculate the final opacity value ------- 39 from a given plume data signal. During the data reduction process (Section 2.4.2 of this Method) the values of Op (which was calculated from the applicable 0 and 0 values) and S p pc o shall be documented along with the applicable parameters used in performing the running average. The date and time that each lidar data signal was obtained, its respective assigned control number, its magnetic tape file address and the tape file address of the respective reference measurement shall also be recorded for each final opacity calculation. The identity of each criterion used in the data analysis and the identity of any opacity values rejected [Section 2.4.3 of this Method] shall be recorded for each applicable opacity value. 3. Lidar calibration and operational error. The lidar shall be calibrated in the field to ensure the accuracy of the opacity data as measured. The calibration also verifies the opera- tional integrity of the optical receiver and the analog/digi- tal electronics. 3.1. Calibration requirements. In there are six parameters that This is accomplished by means the calibration of the lidar shall be directly verified, of two techniques. Technique 1 - Use of neutral density optical filters to simulate opacity with a He~Ne laser or a light-emitting diode as the optical (red light for a lidar containing a ruby laser) source. The narrow band filter (6943 Angstroms peak) shall be removed [Section 4 of this Method] from its position in front of the photo-multiplier (PMT) detector. Neutral density filters of nominal opacities of 0, 20, 40, 60 and 80% shall be used. The recommended test configuration is depicted in Figure AM1-11. ------- PMT Entrance Window Completely Covered U Lidar Receiver Photomultipller Detector (a) Zero-Signal Level Test He-Ne Laser or Light-Emitting Diode (Light Source) light path Lidar Receiver Photomultiplier Detector (b) Clear-Air or 0% Opacity Test Neutral-density optical filter / He-Ne Laser or Light-Emitting Diode (Light Source) light path Lidar Receiver Photomultipl ier Detector (c) Optical Filter Test (simulated opacity values) *Tests shall be performed with no ambient or stray light reaching the detector. Figure AMI-11. Recommended Test Configuration for Technique 1. ------- 41 Technique 2 - Use of an optical generator (built-in calibra- tion mechanism) that contains a highly-controlled light-emit- ting-diode (red light for a lidar containing a ruby laser) which, by injecting an optical signal into the lidar receiver immediately ahead of the PMT detector, simulates an actual lidar return from a given atmospheric path through a plume or in clear air. The optical generator shall simulate opti- cal signals representing clear air or 0% opacity, 20, 40, 60 and 80% opacities (nominal). An optional 10% simulated optical signal may also be included. The six parameters are the following: 1. The zero-signal level (receiver signal with no optical signal from the source present) shall be verified for proper digitizer adjustment or video offset on the oper- ator's oscilloscope display, and there shall be no spur- ious noise in this signal. 2. The opacity value of 0% shall be verified from the light source through the PMT detector, data processing elec- tronics and as calculated by computer. 3. A minimum of 4 opacity (>0%) values shall also be veri- fied through the lidar detector, data processing elec- tronics and as calculated by computer. The neutral density filters and the optical generator shall be calibrated for actual opacity once per month while in use. The calibrated opacity value of each filter or each optical generator simulated opacity value shall be recorded in the lidar log book [Figure AM1-3]. The filters shall be calibrated within a laboratory accuracy of + 2% or better. The optical generator shall be calibrated with an accuracy of + 1% or better. ------- 42 The calibration of the lidar is conducted for each of the two video signal channels (linear and logarithmic [if so equipped] channels) as they are used for data collection. The calibration shall be performed for each emissions source tested. For one given source-under-test the calibration shall be performed once (minimum) every 4-hour period. 3.2. Calibration procedures. With the entire lidar receiver and analog/digital electronics turned on and adjusted for normal operating performance, the following procedure shall be used for Technique 1. 3.2.1 Procedure for Technique 1. This test shall be performed with no ambient or stray light reaching the PMT detector. The zero-signal level shall be measured and recorded (on magnetic tape) as indicated in Figure AMl-ll(a). The simu- lated clear-air or 0% opacity value shall be tested by using the selected light source depicted in Figure AMl-ll(b). (A laser beam may have to be optically attenuated so as not to saturate the PMT detector). This signal level shall be mea- sured and recorded on magnetic tape. The opacity value is calculated by taking 2 100-nanosecond pick intervals about 1 microsecond apart in time, and using Equation AM1-2 setting the ratio R /R, = 1. This calculated value shall be recorded n f in the lidar log book. The simulated clear-air signal level is also employed in the optical test using the neutral density filters. Using the test configuration in Figure AMl-ll(c), each neutral density filter is separately placed into the light path from the light source to the PMT detector. The signal level is mea- sured and recorded manetic tape. The opacity value for each filter is calculated by taking the signal level for that respective filter (If), dividing it by the 0% opacity signal level (In) and performing the remainder of the calculation ------- 43 by Equation AM1-2 with Rn/Rf = 1. The calculated opacity value for each filter shall be recorded in the lidar log book. The file address for each filter-signal recorded on magnetic tape shall be recorded in the lidar log book. 3.2.2 Procedure for Technique 2. With the entire lidar receiver and analog/digital electronics turned on and adjusted for normal operating performance, the optical generator is turned on and the signal, uncorrected for 1/R2, selected with no plume spike signal or opacity value selected. This is a simulated clear-air atmospheric return signal is displayed on the system's oscilloscope display. The lidar operator then makes any fine adjustments that may be necessary which may include detector high voltage and input/offset values on the digitizer (analog-to-digital signal converter). This signal is then switched to the 1/R2 corrected signal. With the digital tape recorder turned on, the magnetic data tape properly marked and installed, and the computer program/ selectable parameter values read into the computer memory via magnetic tape or keyboard, the system is ready for the calibration verification test. Either the linear or the logarithmic video channel is selected that will be used to measured stationary source emissions. (The linear channel is used to measure plume opacity values less than 50 to 60% while the logarithimic channel (optional) is used for opacity values greater than 50% and where large video signal dynamic range is required such as in an urban area where pollutant loading in the local atmosphere is quite high.) If both channels are expected to be used, then both shall be checked for calibration. The opacity values for 0% and the other four or five values are selected one at a time in any order. The computer receives the simulated return signal data, records each signal on magnetic tape and calculates the respective opacity value ------- 44 (this measurement/calculation shall be performed at least 3 times for each selected opacity value). Each of the opacity values from the optical generator shall be verified, while the order is not important, for the respective video channel to be used. The calibrated optical generator opacity value for each selec- tion is recorded in the lidar logbook in the Lidar Function Verification Section [Figure AM1-3]. The average of the three opacity values calculated by the computer for each selection is also recorded in the logbook. The file location or address on the magnetic data tape for each respective simulated video signal is also recorded in the logbook. 3.3. Calibration error. If after the calibration procedure has been completed the lidar measured opacity value is not within +3% (full scale) of the actual calibration value (opacity value of the neutral density filters or the optical gener- ator) as follows: Linear Channel - + 3% over the opacity range of 0% through 60% Logarithmic Channel - + 3% over the opacity range of 20% (Optional) through 80% then the lidar shall be considered out of calibration. Remed- ial action shall be taken to correct any equipment or computer problems that may be present. The calibration test shall be performed again. This process shall be repeated until the above conditions are fulfilled. 4. Specification for Basic Lidar Apparatus. The performance/ design specification for a basic lidar system is provided in this section. This specification is directly addressed to a lidar system using a ruby laser as the optical transmitter. ------- 45 The essential components of a basic lidar system are depicted in Figure AM1-12, the functional block diagram. The types of components given and identified for this system have been proven in the EPA Omega-1 Lidar. Other components may cer- tainly be used in fabricating a lidar system if they meet or exceed the relevant performance characteristics of the compo- nents identified in this section. The components of the lidar instrumentation are identified as follows: a) Q-switched ruby laser - The laser shall have a minimum output energy of 1 joule per pulse with a pulse duration (length) of less than 35 nanoseconds (full width, half power). It shall be capable of being fired at a sustained rate of 1 pulse every 10 seconds (6 pulses/minute). The beam divergence (beam spread with range) angle shall be 1.5 milliradians (full angle, half energy) maximum, which may necessitate the use of an up-col1imating tele- scope on the front of the laser. Inherent laser beam divergence is a function of the ruby rod quality, the higher the quality the smaller the divergence angle. - Holobeam Series 300 and 600 laser, equivalent perfor- mance or better meeting the specifications given above. (b) Reflective Telescope - The telescope shall be a cassegrain reflective instrument with a minimum primary mirror diameter of 20 cm (8 in). A standard amateur astronom- ical telescope performs efficiently at the ruby (red light) wavelength. - Celestron 8 inch reflective telescope, equivalent performance or better. (c) Aiming Telescope - This device is a standard rifle scope with an optical power or magnification of 10 or 12. ------- Mam Power Supply Transmitted Light Pulse Q-Switch Power Supply Sw itched W ater Coo ler Ruby Laser Laser Controls Backscatter Return Signal Reflective Telescope PMT Detector Logarithm ic (C ass eg rain) (b) Logarithmic Video Signal iming Telescope F ast Trans lent Recorder or Digitizer PMT Detector H igh Voltage Interface Bus Power Supply Digital Computer D ig it a 11 /R 2 Correction Opacity Calculation (Variable) (d) (I) Interface Bu Pr inter M agnetic Tape (paper output) Figure AM1-12 Functional Block Diagram of a Basic Lidar System 4* <7> ------- 47 (d) PMT Detector Power Supply - This is a standard PMT high- voltage power supply variable over the range from 1.0 to about 3.0 KVDC. (e) Photomultiplier Detector - The PMT used in the EPA Omega-1 Lidar is of special design to eliminate the problems of afterpulsing or signal induced noise encountered in standard off-the-shelf detectors. This detector has 8 dynodes and a gain of 10s at less than 2.5 KVDC. It has a low impedance photo cathode. The measured quantum efficiency is approximately 5% of the ruby wavelength of 6943 Angstroms. Its spectral response approximates S-20. - ITT PMT Detector No. 4084 fabricated to EPA specifi- cations, equivalent performance or better. (f) Narrow Band Optical Filter - This filter keeps all opti- cal radiation that enters the receiver from reaching the PMT detector with the exception of the extremely narrow band which contains the ruby wavelength. Its center wavelength is 6943+2, -0 Angstroms at 75°F. Its half height bandwidth is less than 14 Angstroms. The peak optical transmittance of this two-cavity, type 2 filter is approximately 51%. - Dell Optics Co. Laser Optical Interference Filter Part Number H-10, equivalent performance or better. (g) Logarithmic Channel Amplifier - This amplifier basically deamplifies strong return signals, gain <1, while it greatly amplifies weak return signals, gain >> 1. Its bandwidth is 30 MHz with a rise time of 11.7 nanoseconds. Its linearity is less than + 1.0 dB. Its slope is nomin- ally 25 millivolts/dB with an overall gain of 100 dB. ------- 48 - Aertech Industries Logarithmic Video Amplifier Model LDN-1000-1, equivalent performance or better. (h) Fast Transient Recorder (Digitizer) - This instrument converts the analog backscatter (or logarithmic video) signal to a computer-compatible digital form. The out- puts are to the computer and the oscilloscope display which is continuous due to refresh memory. - Biomation 8100 Waveform Recorder, equivalent perfor- mance or better. (i) Digital Computer - The computer carries out the required data handling, equipment address assignments and data transfer to the magnetic tape recorder in addition to performing the plume opacity and other calculations. - Hewlett Packard 9825A Programmable Calculator, Digital Equipment Corp. PDP-11-03 or Data General Nova 3 compu- ters, equivalent performance or better. (j) Magnetic Tape Drive/Formatter - This compound instrument records and plays back the lidar backscatter signal data. It can be 800 or 1600 bits per inch with a nominal read/ write speed of 25 inches-per-second. Tape format shall be industry compatible. - Kennedy Tape Drive Model 9800 interfaced to a Dylon Corp. 1015 A Formatter, equivalent performance or better. (k) Oscilloscope Display - The scope provides a continual monitor of the lidar backscatter signals for the lidar operator. - Tektronix Model 475 oscilloscope, equivalent or better. ------- 49 (1) Digital Clock - The digital clock provides time to the nearest second to the computer for each lidar backscatter signal. The time on the clock is set either manually or by the computer. - Hewlett Packard Digital Clock Model 59309A, equivalent performance or better. (m) Printer - Any standard line printer can be used which can be directly interfaced to the computer chosen. In the case of the Hewlett Packard 9825A a thermal printer is an integral part of the machine and another printer is not required. As depicted in Figure AMI-12, the laser and the telescope are mounted biaxially on a steerable mount or pedestal along with the PMT detector and aiming telescope. The minimum convergence distance between the laser beam and the telescope's instantaneous field-of-view shall be 50 meters. Extra care must be taken not to route the video signal cables (coaxial cable 5On impedence) adjacent to power, high voltage and remote control wiring to eliminate electromagnetic interference. The AC power for the electronics must be regulated and isolated from the 220VAC power lines used for the laser. The list of references in Reference 5.1 in addition to the text of this reference provides much information on the theory, design fabrication and use of ruby lidars. Each lidar system whether it employs a ruby laser or another laser as the optical transmitter, must be subjected to performance and evaluation tests, and properly qualified in order to be used under this method. Several tests of this type are described in Reference 5.1. The computer software necessary to fulfill the requirements of this method can be readily developed from the contents of Section 2 of this Method. ------- 50 5. References 5.1 The Use of Lidar for Emissions Source Opacity Determination, U.S. Environmental Protection Agency, National Enforcement Investigations Center, Denver, CO, EPA-330/ 1-79-003, Arthur W. Dybdahl, current edition. 5.2 Field Evaluation of Mobile Lidar for the Measurement of Smoke Plume Opacity, U.S. Environmental Protection Agency, National Enforcement Investigations Center, Denver, CO, EPA/NEIC-TS-128, February 1976. 5.3 Remote Measurement of Smoke Plume Transmittance Using Lidar, C. S. Cook, G. W. Bethke, W. D. Conner (EPA/RTP). Applied Optics 11, pg 1742, August 1972. 5.4 Lidar Studies of Stack Plumes in Rural and Urban Environ- ments, EPA-650/4-73-002, October 1973. Laser Safety References: 5.5 American National Standard for the Safe Use of Lasers ANSI Z 136.1-176, 8 March 1976. 5.6 U.S. Army Technical Manual TB MED 279, Control of Hazards to Health from Laser Radiation, February 1969. 5.7 Laser Institute of America Laser Safety Manual, 4th Edition 5.8 U.S. Department of Health, Education and Welfare, Regula- tions for the Administration and Enforcement of the Radiation Control for Health and Safety Act of 1968, January 1976. 5.9 Laser Safety Handbook, Aiex Mallow, Leon Chabot, Van Nostrand Reinhold Co., 1978. ------- |