EPA-650/2-73-027 October 1973 ENVIRONMENTAL PROTECTION TECHNOLOGY SERIES I 1 ------- EPA-650/2-73-027 AN INSTRUMENT FOR SIMULTANEOUS MONITORING NOx AND S02 IN STATIONARY SOURCES by Huel C. Tucker and Joseph Cheng Monsanto Research Corporation Dayton Laboratory Dayton, Ohio 45407 Contract No. 68-02-0554 Program Element No. 1A1010 EPA Project Officer: Fred Jaye Chemistry and Physics Laboratory National Environmental Research Center Research Triangle Park, N.C. 27711 Prepared for OFFICE OF RESEARCH AND DEVELOPMENT U.S. ENVIRONMENTAL PROTECTION AGENCY WASHINGTON, D.C. 20460 October 1973 ------- This report has been reviewed by the Environmental Protection Agency and approved for publication. Approval does not signify that the contents necessarily reflect the views and policies of the Agency, nor does mention of trade names or commercial products constitute endorsement or recommendation for use. ii ------- ABSTRACT A Monsanto Model 3^09 Chemilumlnescent Ambient Air Monitor was converted to a two-channel configuration for simulta- neously monitoring NOX and S02 in stack gases. Channel separation was obtained by means of narrow-band optical filters. The analog sample-hold part of the automatic zero circuit was replaced by a digital memory circuit. A permeation dryer was included to remove water from the sample. The atomic oxygen source was changed to an ozone generator-thermal decomposition configuration, which failed to provide sufficient oxygen for S02 or N02 analysis. A microwave oxygen generator provided adequate quantities of oxygen but was rejected because of its unreliability. iii ------- CONTENTS Page ABSTRACT ill CONTENTS iv FIGURES v TABLES vi ACKNOWLEDGEMENTS vii SECTIONS I. INTRODUCTION 1 II. RESULTS OF EXPERIMENTAL WORK 3 III- CONCLUSIONS 5 IV. TECHNICAL DISCUSSION 7 Design 7 Sampling System 9 Oxygen Generators 11 Reaction Chamber 12 Pneumatic Circuit 12 Electronic Modifications 16 Programmer 16 Amplifier 20 Connecting Wiring 22 RESULTS OF EXPERIMENTAL WORK 22 Permeation Dryer 22 Oxygen Generator 22 APPENDIX 33 iv ------- FIGURES No. Page 1 Sample Probe Schematic 10 2 Ozone Generator Photograph 13 3 Decomposition Chamber Photograph 13 4 Reaction Cell Photograph 14 5 Detector Assembly 14 6 Pneumatic Schematic 15 7 Clark Circuit Schematic 17 8 Logic Circuit Schematic 18 9 Solenoid Driver Circuit Schematic 19 10 Amplifier Circuit Schematic 21 11 Response to NO and N02 with Ozone Generator 24 12 Schematic for Microwave Excitation 29 13 Response to S02 With Microwave Excitation 30 14 Front Panel Layout 31 ------- TABLES No. Page 1 Response to NO, S02 With Ozone 25 2 NO Response With Ozone 26 3 Response to N02 27 4 Response to S02 With Microwave 28 vi ------- ACKNOWLEDGEMENTS Significant contributions were made to this project by G. W. Wooten, Dr. A. D. Snyder, C. M. Ellas, and D. J. David of Monsanto Research Corporation and Dr. F. J. Jaye of EPA. vii ------- SECTION I INTRODUCTION Three major air pollutants are NO, N02 and S02. Each may be monitored with available commercial instruments. However, no one instrument is capable of analyzing for all three simultaneously. Simultaneous monitoring of all three pollutants requires two or three instruments depending on whether one of them has the capability to directly monitor NOX, which is the sum of the NO and N02 present. Since NO and N02 are usually lumped as NOX, the Environmental Pro- tection Agency (EPA) deemed it to be desirable to develop an instrument to simultaneously monitor S02 and NOX for coal- fired stack emission monitoring. Previous work in the area of chemiluminescence indicated that the reaction of nitric oxide with ozone could be used in relatively straightforward equipment to measure ambient air levels of NO and N02. There are a number of N02-to-NO converters in use with such ambient air equipment. Work in the stationary source area was somewhat complicated by the presence of side reactions that can create or destroy NO or N02 in these converters. The work by G. W. Wooten using microwave-produced atomic oxygen instead of ozone indicated that the reaction N02 + 0 >• NO + 02 could be used instead of an external converter to monitor total NOX. Analogous S02 chemiluminescent reactions could also be used to measure S02 at the same time. ------- The potential desirability of a dual S02-NOX monitor for certain stationary sources prompted this contract effort. The modification of an experimental chemiluminescence in- strument which was designed for low-level NOX analysis was decided upon as the most expedient approach to the design of such a dual analyzer. The design goal was to develop an instrument for continuously and simultaneously monitoring NOX and S02 at high levels at remote locations with minimal attention from the operator. Evaluation of the original instrument showed that the analog sample hold circuits used for zero drift correction were not adequate over long time periods. With the recent advances in digital integrated circuitry in terms of functions and cost, we decided to rework the auto-zero circuitry to use digital components. Concurrently with the initiation of this program, a sample drying system based on selective water permeation became commercially available. Since removal of condensable water was required for the proper operation of the modified in- strument, we undertook to incorporate this new dryer into the sampling system. ------- SECTION II RESULTS OF EXPERIMENTAL WORK The ability of the permeation dryer to selectively remove water from a sample stream was confirmed. However, no effort was made to establish the operating parameters and their limitations. The automatic zero circuit employing a digital memory was successfully demonstrated. This technique extends the usefulness of all-electronic automatic zero circuitry to include applications requiring memory retention times measured in days or even weeks as opposed to the 10 to 15 minute maximum memory retention time of analog memories. All attempts to simultaneously measure NOX and S02 using an ozone generator with a thermal decomposition chamber to generate atomic oxygen failed. The response of the instrument to NO was adequate and relatively easy to achieve. The response to N02 was inadequate for a measuring circuit having no background noise compensation. N02 response was difficult to obtain and required that all parameters be optimized. No response to S02 was observed when generating oxygen by this method. It was not established if the response existed or if it was merely hidden in the noise of the system. It was established that the response to NO was destroyed by the Introduction of small amounts of S02. The previous work where response to S02 and N02 was reported was confirmed by a test using a microwave generator and a cavity to generate atomic oxygen. Again, the response to NO was easy to obtain. An adequate response to S02 was obtained only by careful and methodical optimization of all ------- parameters of the reaction chamber. No effort was made to explain the effects of the various parameters. ------- SECTION III CONCLUSIONS A chemilumlnescence instrument based on atomic oxygen to simultaneously monitor stationary emission sources for NOX and S02 is impractical with the present state of the art. Further clarification of the requirements of the chem- iluminescence reaction of atomic oxygen with various com- pounds such as N02 and S02 is required before any serious consideration can be given to further development of such an instrument. Although the instrument can be made to function adequately over a short time period by using microwave excitation to generate atomic oxygen, the mechanical instability of the microwave cavity-to-oxygen line system, the projected short life of the microwave tube, and the uncertainties related to maintaining a plasma discharge over a long period of time make that approach unacceptable for a continuous monitor. Either the ozone generator/thermal decomposition method for making atomic oxygen failed to produce sufficient oxygen, or an unknown reaction between S02 and another compound present had a quenching effect on the chemiluminescence reaction. The saturation effect observed with N02 indicates that the NO measurement can be made in the absence of S02 by using a second photomultiplier to compensate for background variations by limiting the flow of the sample into the reaction chamber to keep it within the linear range. The successful application of a digital memory to an automatic zero circuit represents a significant advance in the state ------- of the art pertaining to programmed field instruments. It can replace high maintenance servo systems in almost any application where the inherent drift of an analog memory is unacceptable. It may also be useful as a long-term sample- and-hold amplifier. ------- SECTION IV TECHNICAL DISCUSSION DESIGN The overall design goals were outlined by parts A and C of the following Scope of Work statement, included as Exhibit A in the contract: "A. The contractor shall modify the existing Monsanto Model 3409 ambient air NOX monitor to simultaneously measure stationary source S02 and NOX emissions in the following ranges: S02 - 0-200 ppm, 0-500, 0-1000, 0-3000 ppm NOV - 0-200 ppm, 0-500, 0-1000, 0-3000 ppm A. The unit shall have equivalent responses to NO and N02 or mixtures thereof without the use of a con- version device. B. Tasks required in this conversion include, but may not be limited to: 1. Replacement of differential signal processing circuits and replacement with two independent channels including independent automatic zero correction circuits and manual span adjustment for both channels. 2. Removal of discharge atomic oxygen generator and replacement with ultraviolet ozone generator with small thermal decomposition furnace downstream (1" diameter, 2" long, 800°C) or microwave genera- tor. 3. Replacement of reaction chamber with two port chamber for simultaneous S02 and NOX measurements. 4. Replacement of optical filter on second photo- multiplier with filter centered at 3510 A, % width 50 A. ------- 5. The unit shall retain the internal automatic zero correction and calibration functions. C. The contractor shall provide a sampling system suitable for operation at a coal-fired steam electrical power plant. This sytem shall include particulate removal and other sample preconditioning as required by the covered prototype. D. The contractor shall calibrate the instrument for S02, NO and N02 in the specified ranges and establish inter- ference ratios for 500 ppm CO, 3% 02, 100 ppm C2H6, IQ% H20, 15% C02, 100 ppm N20, and 500 ppm NH3. E. The converted monitor and sample system shall be deliver- able. The contractor shall document and blueprint all changes. The contractor shall provide fifty (50) copies of a design and operation manual and a set of engineer- ing drawings which will be sufficient to enable the manufacture of items of equipment furnished under this contract (other than components or items of standard commercial design or items fabricated heretofore) by a firm skilled in the art of manufacturing items of equip- ment of the general type and character of the items or equipment furnished under this contract or a set of flow sheets and engineering drawings which will be sufficient to enable performance of any process developed under this contract by a firm skilled in the art of practicing processes of the general type and character of good process. Such set or sets of drawings and flow sheets shall be reproducible copies incorporating all changes made in the equipment or process in the form in which it was delivered to the Government." Parts B, D and E in the above statement pertain to more specific details to be included in the overall design. Design decisions were based on this statement but were influenced by additional information supplied by the contract monitor. The design goal of up to 3000 ppm of each NOX and S02 was modified to 3500 ppm sum of NOY and S02 on the assurance of J\, the monitor that that capability would be sufficient. ------- Sampling System The requirements of the sampling system were set forth in parts C and D of the Scope of Work statement, above. Further clarification by the contract monitor established the need for removing much of the water from the sample at the point of exit from the stack to prevent condensation as the water- laden sample gas cooled enroute to the analyzer. Because of the highly reactive nature of N02 and S02, any condensation in the sample system would have a drastic effect on these analyses. At the recommendation of the monitor, a permeation dryer was obtained and tested for its effects on S02 and N02. Simple test procedures were used where 3502 ppm S02 and 11350 ppm N02 were passed through the dryer and analyzed to determine any changes in the concentrations. There was no detectable change in S02 and no more than 2% change in the concentration of N02 where the water content of the drying gas was held low. However, it was found that where the water content of the drying air was high, the effect on the concentrations of the S02 and N02 at the exit of the dryer was drastic in that much of the S02 or N02 was not passed by the dryer. A sampling probe was designed utilizing a Model PD-500-40 permeation dryer purchased from Permapure Products, Inc. With reference to the schematic shown in Figure 1, the sample gas is pulled through the permeation tube via a coarse filter at the end of a 6 ft long, 1/2 inch stainless steel pipe designed to be inserted into a stack. After exiting the dryer, the dry gas is routed through flexible tubing to the analyzer. Dry air, obtained from a bottled supply or through a dryer, is pulled through the vacuum compartment via a valve, VI, by pump, PI, and discharged through a solenoid valve, SV2, to ------- Dry Air Inlet Filter 1/2",6'Long S.S. Pipe/ Flexible Tubing to Analyzer Wooden Enclosure Vent Figure 1. Sample Probe Schematic ------- atmosphere. During the "zero" and "standard" cycles, where a bottled zero or standard gas is being accepted by the analyzer, the tubing is blocked and SV2 closes, SV1 and SV3 open, routing the effluent from pump, PI, in a reverse direction through the permeation tube, probe, and filter. This procedure should remove some particles collected on the filter, thereby minimizing the maintenance required for cleaning it. Oxygen Generator The design of the oxygen generator was the most critical phase of the project. The two approaches that were con- sidered feasible were (1) the direct disassociation of 02 using microwave power and (2) the indirect method of generat- ing ozone and thermally decomposing it. A survey of the ozone generators available revealed that only the first method could produce sufficient concentrations of atomic oxygen to accommodate the high concentrations of pollutants expected. However, agreement was reached with the contract monitor that, considering the present state of the art, the solution of the problems to be incurred in the use of a microwave generator in an instrument of this type where continuous, unattended operation was required was beyond the scope of this project. An investigation revealed that commercial ozone generators were available. Previous work reported that linear results could be obtained by the ozone/thermal decomposition methods to well over 100 ppm NO in N2. The reported signal level X at 100 ppm was adequate to provide a readout for the instru- ment. Therefore, it was decided to design the oxygen generator using the ozone/thermal decomposition method, and 11 ------- to dilute the sample gas with air to reduce the pollutant concentration to an acceptable level. An Orec Model 03V5-OM ozonator was purchased and a thermal decomposition chamber designed to provide atomic oxygen. The generator is shown in Figures 2 and 3. Reaction Chamber The reaction chamber was designed to mix the sample stream with a stream containing the atomic oxygen. It provides observation ports at opposite ends for two photomultipliers. There is a relatively long path for the mixed stream to exit to the vacuum pump as shown in Figures 4 and 6. The chamber is housed in a light-tight box as shown in Figure 5. The flow rates of both sample and oxygen/ozone are controlled by stainless steel needle valves to provide the proper mixing, A by-pass valve provides control of the pressure in the chamber. Pneumatic Circuit The plumbing for the instrument is shown by the schematic in Figure 6. All fittings, tubing, valves, etc., are stainless steel, glass, quartz, or inert plastic. The pump, P2, pulls the sample and the diluent air past the inlet to the reactor. The mix of air and sample is controlled by valves V5 and V2, respectively. The flow of the diluted sample into the vacuum reaction chamber is controlled by V6. Valves V8 and V9 are manipulated to obtain optimum flow of oxygen through the ozone generator and to maintain near atmospheric pressure in the ozone generator. Atomic oxygen is generated as the ozone passes through the heated quartz tube. The chem- iluminescence occurs within the inner tube as the two gas 12 ------- Figure 2. Ozone Generator Figure 3. Thermal Decomposition Chamber 13 ------- Figure Reaction Chamber Figure 5. Detector Assembly ------- VJ1 Air Inlet— V-5 Standard_ Gas Inlet Sample Inlet Zero Gas Inlet - FI-2 PI (\ V-6 Fl-l Optical Filter-1 I PMT-1 Reaction Chamber Optical Filter-2 -^— Vent iip- v-9 I Exhaust V-8 0- Supply Figure 6. Pneumatic Schematic ------- streams mix. The mixture remains within the field of both photomultipliers as it passes from the point of entry at the center through the inner tube to either end and back to the exit point at the center through the outer tube. The pressure can be controlled by the by-pass around the vacuum pump, P3. Electronic Modification The electronic circuit modifications were divided into three parts. First, the programmer modifications required some circuit changes but no major design. Second, the amplifier circuits required a complete redesign. Third, the connecting wiring required partial redesign. Programmer - The programmer consists of a clock circuit (Figure 7), a logic circuit (Figure 8), and a solenoid driver circuit (Figure 9). The clock consists of a 100 kHz oscil- lator and a five-decade counter circuit for generating a one- second clock pulse. A second counter circuit is programmable for a divide by 1-99 counter to establish the basic time unit for the pin board programmer on the front panel. As wired, the basic time unit is 60 seconds. Thus, a function programmed to remain on 15 time units will actually have a 900 second duration. The programmer is designed to reset itself at a programmable time. The auto zero cycle can be programmed to occur at any time within the interval of the total cycle and to have any desired duration. The electronic auto zero function is on only during the last time unit (60 seconds) of the auto zero cycle. The standard gas cycle can be programmed to occur at the beginning of each.Nth cycle where N is a number greater than 1. A typical program would have a auto zero cycle of 10 time units (600 sec.) occurring 16 ------- CLOCK IOAIO 1 CLOCK BOARD 2 Vcc Ground I.CNo. Pin No. Pin No. 1 7 8 9 10 11 12 13 I 14 14 14 14 14 14 14 SP MO A SN7440N SN7440N SN7440N SN7400N SN7404N SN 7402 N SN7400N Figure 7. Clark Circuit Schematic ------- co SN 74.04 N SN74.WN SN 74LM N SN74IBN SN74.10N SN7400N SN7400N SN7CON Figure 8. Logic Circuit Schematic ------- •H4V H VO +24V Figure 9. Solenoid Driver Circuit Schematic ------- each cycle of 2^0 time units (4 hours) with a 10 time unit (600 sec.) standard cycle each 6 cycles (24 hours). Amplifier - The amplifier circuit was completely redesigned and new printed circuits made. Identical channels were provided for each of two photomultipliers. Because the intervals between the auto zero cycles were long, a digital memory circuit was incorporated to prevent drift normally associated with capacitor leakage in analog memory circuits. The circuit is shown in Figure 10. The amplifier consists of three stages. The first is a LM308A operational amplifier with adjustable gain and a front panel zero adjustment. The second stage is a LM301A with a front panel gain adjust and a zero control input from the auto zero circuit. The third stage is a LM201A amplifier with a switchable gain on the front panel and a zero adjustment on the printed circuit board. The output of the second stage also goes to the input of a LM201A acting as a comparator such that its output always has the opposite polarity of the second-stage amplifier and is much greater in magnitude. Therefore, when the second- stage amplifier output is not zero, the output of the com- parator is large. The comparator drives a Schmitt trigger (SNT^ISN) which in turn drives a logic circuit controlling a two-stage, binary up-down counter whose digital output drives a digital-to-analog (D-A) converter (DAC372-8). Upon being enabled by a signal from the programmer or a switch on the front panel, a 100 pulse per second clock signal is routed through the logic circuit to the up-down counter, which counts up or down depending on the polarity of the comparator. The analog output of the D-A converter is fed to one input of the second stage LM301A such that the LM301A output is always driven toward zero. The system responds much as a slightly oscillatory servo system in that the LM301A output 20 ------- »ONT PANEL L C 470 < Figure 10. Amplifier Circuit Schematic ------- oscillates ± one bit around zero. By making one bit in- significantly small, the output of the LM301A can be forced to zero for all practical purposes. Upon removal of the auto zero signal, the counter is disabled, but it retains its digital output. Therefore, the analog zeroing signal from the D-A converter is maintained until the circuit is again activated. Connecting Wiring - The connecting wiring was modified as required. A new harness was made for the front and rear interconnections and the mother board was cut and patched where necessary. The new wiring diagram is included in Figure 10. EXPERIMENTAL WORK Permeation Dryer The ability of the Model PD-500-40 Perma Pure dryer to pass S02 and N02 unchanged was determined. Water vapor, N02 and S02 at various known levels of concentration were introduced into the sample inlet of the dryer while the outlet was monitored to detect losses in transient. It was found that significant losses were incurred unless the drying air in the outer shell was much dryer than the sample. Almost total loss of N02 was experienced under wet air conditions. How- ever, no detectable change in S02 and only a small 1-2% change in N02 was found when dry air was used. Oxygen Generator The Orec Model 03V5-OM Ozonator failed to operate satisfac- torily within the recommended input current range of 0-150 mA 22 ------- but did function at 175-250 mA. Sustained operation at that level produced no detectable ill effects. It was found that the pressure affected the operation of the unit. Optimum performances seemed to occur at 0-2 psig. The modification of the monitor was delayed for some time awaiting delivery of the optical filters. The 6500 ± 20& filter was received and installed in mid-November 1972. A test was run to test the response to NO with the results shown in Figure 11. No effort was made to calibrate the instrument at that time since the 3400& filter had not been delivered. In view of the considerable delay in the delivery of the 3400& filter, a Fish-Schurman UG-1 filter was ordered and received late in January 1973. After demonstrating that linear results could be produced by microwave excitation to produce atomic oxygen, an effort was again made to obtain positive results using the ozone/thermal decomposition method. Again, no detectable response was obtained for S02. The response to N02 is shown in Figure 11. Although the UGI filter was inferior to the T-5, Infrared Industries filter,, it was useful in evaluating the performance of the instrument. On February 1, 1973 an effort to detect S02 failed. The gain of the amplifiers was increased to a point where the noise level was detectable and the results tabulated in Table 1 were obtained. As a result of the failure to detect S02, a test was devised using a microwave generator to generate atomic oxygen as shown in Figure 12. The results of tests are shown in Figure 13. 23 ------- 50 Qi 8 CT> TO 30 20 =3 o 10 0 1550 ppm NO, 02 Flow: 200 ccm 1550 ppm NO, 02 Flow: 300 ccm 1550 ppm NO, 02 Flow: 500 ccm 4350 ppm N02, 02 Flow: 200 ccm 4350 ppm N02, 02 Flow: 300 ccm 4350 ppm N02, 02 Flow: 500 ccm 0 5 10 15 Sample Flow Rate, cc/min 20 Figure 11. Response to NO and N02 with Ozone Generator ------- Table 1. RESPONSE TO NO, S02 WITH OZONE Conditions: Reaction Chamber Pressure: Ozone Generator Current: 02 Flow Rate: Decomposition Heater Temperature: Samples: 1550 ppm NO in N2 280 ppm S02 in N2 1050 ppm S02 in N2 1-5 torr 250 mA 200 cc/min 800 - 900°C NOX Filter: S0? Filter: 6500 ± 50) 3^00 ± 50J •J-J-^GJ. . JTUU J. ~j\jn. (S02 filter0transmission is appreciable above 6000A) NOX photomultiplier: EMI 9781-B S02 photomultiplier: RCA Ip 28 NOX gain: S02 gain: 0.50 maximum State of Sample Input Ozone Gen. off No sample Ozone Gen. on No sample Ambient Air NO S02 (280 ppm) S02 (1050 ppm) S02 (1050 ppm) + NO NO S02 (1050 ppm) + NO S02 (1050 ppm) + NO NOX Output, mV (1000 mV fs) ± 5 5 ± 5 30 ± 10 550 ± 10 30 ± 10 30 ± 10 30 ± 10 250 30 ± 10 30 ± 10 Flow Rate of Sample, cc/m 0 0 100 100 100 100 50 ea. 50 50 ea. 50 NO, 5 S02 S02 Output mV ± 15 5 ± 15 20 ± 20 110 ± 20 20 ± 20 20 ± 20 30 ± 20 50 ± 20 30 ± 20 30 ± 20 ------- Table 2. NO RESPONSE WITH OZONE Conditions: Ozone Generator Current: 225 mA Ozone Generator Chamber Pressure; Decomposition Heater Volts: 70 Sample: 1550 ppm NO in N2 Span Gain: 4.00 Ozone Generator Input: 02 Range Switch: 0-3000 ppm 1-5 psig Reaction Chamber Pressure, torr 1.2 1.2 1.2 1.2 1.2 1.3 1.3 1.3 1.3 1.3 1.5 1.5 1.5 1.5 1.5 °2 Flow Rate, cc/min 200 200 200 200 200 300 300 300 300 300 500 500 500 500 500 Sample Flow Rate, cc/min 0 25 50 75 100 0 25 50 75 100 0 25 50 75 100 Output, % (full scale) 0 20 30 40 50 0 23 36 42 53 0 19 28 35 44 26 ------- Table 3- RESPONSE TO N02 Conditions: Ozone Generator Current: 225 mA Ozone Generator Chamber Pressure: Decomposition Heater Volts: 90 Sample - 1*350 ppm N02 in N2 Span Gain - 4.00 Ozone Generator Input - 02 Range Switch 0 - 3000 ppm 1.5 psig Reaction Chamber Pressure, torr 0 1-3 1.3 1.3 1.3 1.3 1.3 1.3 1.5 1.5 1.5 1.5 1.5 1.5 1.5 02 Flow Rate, cc/min 200 200 200 200 200 200 200 300 300 300 300 300 300 300 500 500 500 500 500 500 500 Sample Flow Rate, cc/min 0 5 10 25 50 75 100 0 5 10 25 50 75 100 0 5 10 25 50 75 100 Output, % (full scale) 0 5 7 8 8 8 8 0 6 9 12 12 12 12 0 5 8 13 15 15 16 27 ------- Table 4. RESPONSE TO S02 WITH MICROWAVE Conditions: Sample: 3500 ppm S02 in N2 Span Gain: 3.10 Range: 0-1000 ppm Microwave Power, watts 50 50 50 75 75 75 02 Flow, cc/min 120 153 225 120 153 255 Sample Flow to Reactor, cc/min 0 5 10 20 0 5 10 20 0 5 10 20 0 5 10 20 0 5 10 20 0 5 10 20 Output Reading, % Full Scale 0 18 28 43 0 14 27 44 0 10 22 45 0 21 32 49 0 17 30 50 0 12 25 50 28 ------- ro vo Optical Supply Cylinders, ^Pressure Regulators and Indicators To Vacuum Pump Flow Indicators SO; NO Figure 12. Schematic for Microwave Excitation ------- 50 O) 0 • 120 ccm 02, 50 watts * 153 ccm 02, 50 watts ° 255 ccm 02, 50 watts * 120 ccm 02, 75 watts 153 ccm 02, 75 watts • 255 ccm 02, 75 watts 5 10 15 20 Sample Flow Rate, cc/min Figure 13. Response to S02 With Microwave Excitation 30 ------- Zero Sample Standard aooo Range A3 OA1 A2 *- A1 „„ -* Standard |™ Auto Zero S™ Recycle Cycles of Std. Gas 0123456789 oooooooooo oooooooooo oooooooooo oooooooooo oooooooooo oooooooooo Hundreds 0123456789 oooooooooo oooooooooo oooooooooo oooooooooo oooooooooo oooooooooo Tens 0123456781 oooooooooo ooooo 0,0 o o o oooooooooo oooooooooo oooooooooo oooooooooo Units '0' Range »•» A DA1 \ / *~ A' ~" * 24 v Meter Auto S0 Program Auto Meter Reset NOX Figure Front Control Panel ------- Further experiments results in configurations which produced linear results for S02 and N02 when using microwave excitation. Table 4 shows the effect of oxygen flow rate on the response to both S02 and NO. 32 ------- APPENDIX OPERATING INSTRUCTIONS PROGRAMMING Programming of the instrument is accomplished by means of the recessed pin board matrix switch on the front panel. See Figure 14. The five upper rows of pin holes control the normal cycle program except that the standard cycle occurs only as programmed by the lower row of pins. The third row sets the start of the auto zero cycle and the fourth row sets the end of auto zero cycle. The fifth row sets the time at which the program cycle resets itself and starts over. The first row sets the start of the standard cycle and the second row sets the end. However, the standard cycle only occurs as often as programmed by the lower row of pins. For instance, a pin in the sixth position of the lower row causes the standard cycle to occur once each sixth program cycle. The format for programming is reasonably straightforward. The basic time unit is one minute, and there are three decades on each row such that program cycles can be up to 998 minutes. One peculiarity of programming is that the actual programmed time is one time unit less than the program on the board. Therefore, a one-hour time required a 061 pin setting. 33 ------- AMPLIFIER Each channel amplifier consists of a printed circuit board and various resistors and potentiometers on the front panel. See Figure 10. Each P.O. board contains three integrated circuit analog amplifiers for signal conditioning and one I.C. analog amplifier acting as a comparator which, together with various logic and counting and a digital to analog con- verter, comprises an automatic zero circuit. The first amplifier stage, a high-input impedance LM308A, accepts the current signal from the anode of a photomultiplier tube. The gain this stage is set by a potentiometer on the P.C. board, such that its output is approximately five volts for a 3000 ppm sample. A ten-turn, 5K potentiometer on the front panel provides a suppression current for nulling any background current from the P.M.T. The second stage is a LM301A, the gain of which can be adjusted by the ten-turn, 100K span potentiometer on the front panel. The output of this stage is adjusted to provide five volts at its output, for a 3000 ppm sample. Zero adjustment of this stage comes from the output of a digital-to-analog converter in the automatic zero section. The third stage accepts the signal from the output of the second stage. This signal is amplified by an amount determined by the range switch on the front panel to provide one volt at the output for a sample containing the full scale concentration set on the range switch. Zero adjustment is made by means of a 20K trim potentiometer on the P.C. board. A LM201 amplifier accepts the output of the second stage and generates a large output voltage of the opposite polar- ity. That output is accepted by a Schmitt trigger logic gate ------- which provides a switching action with very small hysteresis. The gate output routes a clock pulse to one of two inputs of an up/down counter, depending on the polarity of the second stage output, through another set of gates which turns the clock pulse train on or off, depending on the state of the command signal which is received through a switch on the front panel. That switch controls the auto zero circuit or connects it to the programmer which turns it on at the end of the auto zero cycle. A push-button switch resets the counters to mid range. The binary output of the counter is accepted by a DAC 372-8, eight bit digital-to-analog converter which supplies a zero signal to the input of the second stage. The action of the total circuit is such that when the auto zero circuit is on, the counter counts in a direction, up or down, to make the D-A converter output change in a direction to cause the second state output to decrease. Upon crossing zero, that output causes the count direction to change so that the circuit oscillates one to two bits around zero. By making the effect of the D-A converter output on the second state output small enough, that oscillation is insignificant. Thus, the circuit operates very much as a slightly oscillatory servo system. When the auto zero is turned off, the counter and D-A converter remain in their last state, thereby re- taining the suppression current to the second amplifier stage input. AUTOMATIC OPERATION In the automatic mode, the programmer transmits logic level signals to the solenoid driver and automatic zero circuits. These signals are derived from the clock counters through decode and latching circuits, such that the automatic zero mode and the standard mode are programmable, as is the 35 ------- frequency of the standard mode and the period of a program cycle. The logic is such that the sample mode is on at all times when neither the zero nor the standard mode is on. In the sample mode, SV4, the sample solenoid, is activated allowing the sample to be drawn through the instrument, see Figure 6. The dryer vent valve, SV2, is also activated, see Figure 1, allowing the drying air in the sample conditioning probe to vent to atmosphere. A part of the gas passing through the instrument is bled through a needle valve, V6, into the reaction chamber where the pollutant reacts with atomic oxygen entering from the opposite side. If necessary, needle valve V5 may be adjusted to provide the required dilution of the sample by entering clean air into the stream. Atomic oxygen is generated by passing molecular oxygen through an ozone generator at near atmosphere pressure and thermally decomposing the ozone as it passes through a heated zone at 800 - 900°C. The flow rate of the ozone/oxygen mixture is set by needle valve V9, which bleeds it into the vacuum chamber ahead of the heated zone. The mixture con- taining ozone and molecular, as well as atomic, oxygen then passes into the reaction chamber where it mixes with the sample and passes in either direction down the inner tube of the reaction chamber to the end and back to the middle where it exits to the vacuum pump, P3. Upon initiation of a zero cycle, the dryer vent solenoid valve and the sample solenoid valve are deactivated, thereby blocking entrance of the sample to the instrument and blocking the exit of the drying air from the permeation dryer in the sample probe. Simultaneously SV1 and SV3 activate allowing an increased flow of drying air to be pumped in a reverse direction through the permeation tube and input filter to 36 ------- vent into the stack. Also SV6 activates allowing a clean dry air flow into the instrument. Needle valve, V4, is adjusted to maintain the same flow through the instrument as was present during the sample mode of operation. During the last minute of the automatic zero mode, the electronic auto zero circuit is activated causing the output of the amplifier to read zero. That level of suppression is then maintained until the next automatic zero mode. Upon initiation of the standard mode, the sequence is identical to that described for the automatic zero mode, except that SV5 is activated, rather than SV6, allowing a standard gas of known concentration to enter the instrument and the electronic automatic zero circuit is not activated. Since the composition of the standard gas is known, span adjustments may be made to compensate for any changes since the last such adjustment, or the changes may be noted or displayed on a recorder for later computations. ELECTRICAL Upon receipt of the instrument, it should be checked for damage in shipping. Check all printed circuit cards and connectors to see that they are seated well. Connect the probe to the monitor by means of the cable with ampherol connectors furnished. Plug all A.C. line plugs into the power distribution panel in the top rear of the monitor. If independent on-off control is desired, use plugs with switches adjacent. Supply 115V, 60 Hertz, to the male receptacle at the rear. Turn on all applicable power switches. 37 ------- PNEUMATIC Connect sample line between the probe and the monitor. Connect zero gas, standard gas, and oxygen lines to the monitor and to the appropriate bottles with regulators. Adjust oxygen flow to 100 CC/min. Connect vacuum pump to outlet at rear of instrument and turn on. ELECTRONIC Adjust Powerstat at rear of monitor for 225 milliamperes ozone generator current. Allow 15 minutes for thermal decomposition heater to stabilize. Set up desired program and reset programmer by pushing the reset switch. INITIAL CALIBRATIONS Provide a calibration sample of known composition and set range switch to appropriate range. Set up a long standard mode period following an automatic zero mode. Allow a sufficient response time (5 to 10 minutes); then switch the meter to read up-scale on Al and adjust the trim potent- iometers on the P.C. boards to 5 volts per 3000 ppm of the pollutants. For example, 1500 ppm should provide 2.5 volts at Al. The meters are calibrated for approximately 10 volts, full scale. Next, switch to A3 and adjust the span potent- iometers on the front panel to provide the correct readings on the meters. Re-program for the desired mode of operation and reset the programmer. 38 ------- |