74-KPM-13 (REPORT NUMBER) AIR POLLUTION EMISSION TEST ALTON BOX BOARD COMPANY (PLANT NAME) JACKSONVILLE. FLORIDA (PLANT ADDRESS) U. S. ENVIRONMENTAL PROTECTION AGENCY Office of Air and Water Programs Office of Air Quality Planning and Standards Emission Standards and Engineering Division Emission Measurement Branch Research Triangle Park, N. C. 27711 ------- EPA REPORT NUMBER 74-KPM-13 ALTON BOX BOARD COMPANY JACKSONVILLE, FLORIDA FINAL REPORT Submitted to Environmental Protection Agency Office of Air Programs Contract No. 68-02-0225 Task No. 24 Submitted by Engineering-Science, Inc. 7903 Westpark Drive McLean, Virginia 22101 December, 1974 ------- TABLE OF CONTENTS Section Title I INTRODUCTION II SUMMARY AND DISCUSSION OF RESULTS III PROCESS DESCRIPTION AND OPERATION A. Process Description B. Process Operation IV SAMPLING AND ANALYTICAL PROCEDURES A. Location of Sampling Points B. Sampling Procedures C. Analytical Procedures 7 11 15 15 15 18 Appendix The Appendices are omitted from this copy ------- Table II-l II-2 III-l LIST OF TABLES Title Particulate Emission Summary Particulate Emission Summary (Metric System) Page 3 4 Summary of Calculations of Equivalent Pulp Production Rate 14 Number III-l III-2 IV-1 IV-2 LIST OF FIGURES Title The Kraft Pulping Process Recovery Furnace and Precipitator at the Alton Box Board Mill in Jacksonville, Florida General Layout, Alton Box Board, Jacksonville, Florida Alton Box Board, Jacksonville, Florida, Stack Test Ports Page 8 10 16 17 ii ------- I INTRODUCTION Under Section 111 of the Clean Air Act of 1970, as amended, the Environmental Protection Agency is charged with the establishment of performance standards for new stationary sources which may contribute significantly to air pollution. A performance standard is based on the best emission reduction systems which have been shown to be technically and economically feasible. In order to set realistic performance standards, accurate data on pollutant emissions is normally gathered from the stationary source category under consideration. The No. 9 black liquor recovery boiler at Alton Box Board Company in Jack- sonville, Florida, was designated as a possible well-controlled stationary source in the Kraft pulp industry and was thereby selected by the Office of Air Quality Planning and Standards for an emission testing program. Tests were conducted by Engineering Science, Inc. personnel during December 10-13, 1973. This plant processes about 650 tons/day of Kraft pulp. The No. 9 recovery furnace is designed to burn approximately 93,750 pounds/hour of black liquor solids. Air pollution control for this furnace consists of an electrostatic precipitator. Samples were collected after the control device to determine filterable and total particulate emissions. Sampling was scheduled during the boiler soot blowing in order to obtain maximum emission levels. The test team and EPA are indebted to Mr. John Jones of the Alton Box Board Company for his cooperation in the sampling program. ------- II SUMMARY AND DISCUSSION OF RESULTS The three particulate sampling runs, which comprised the test pro- gram, were successfully completed during periods of estimated peak emissions. The tests were conducted in the outlet stack of the electro- static precipitator which controls the emissions from the black liquor recovery boiler. It was assumed that maximum emission levels are reached during boiler tube soot blowing. This condition is part of normal boiler operation and occurs at intervals of approximately eight hours. Each soot-blowing cycle takes approximately three hours. Continuous communication with plant-operating personnel ensured that all testing was conducted during periods of soot blowing. A summary of the test results is offered in Tables II-l and II-2. The average total particulate emission rate established by the test results was approximately 117 Ib/hr at an average total particulate concentration of 0.124 gr/scf. The process emission rate for the No. 9 recovery boiler (i.e., the ratio of emission rate to production rate) at the average emission rate of 117 Ib/hr and the average production rate of unbleached air-dried pulp of 19.0 tons/hr was 6.16 Ibs/ton. In an effort to eliminate the possible need for a filter change during the testing, a cyclone separator was installed upstream of the filter holder in the heated sample box. During each test run, condensation collected in the cyclone. This occurred even though the probe heater and sample heater were apparently operating normally. The volumes of liquid collected in the cyclone were as follows: run 1 - 118 ml, run 2 - 31 ml and run 3 - 136 ml. At the 145-foot sampling elevation, the wind was strong and at the first few traverse points, the sample box was suspended about 13 feet ------- TABLE II-l PARTICULATE EMISSION SUMMARY Run Number & Date 2 Volume of Dry Gas Sampled (SCF) Moisture by Volume (%) Orsat Analysis (% by Volume) CO £ Average Stack Temperature Stack Flow Rate (SCFM) Percent Isokinetic 12/11/73 53.66 32.16 10.0 9.1 0 (°F) 249.0 109,190 108.7 12/12/73 53.42 33.70 11.9 7.1 0 239.7 106,826 110.6 12/13/74 55.43 34.39 13.8 4.6 0 241.0 113,807 107.7 Average 54.17 33.42 11.9 6.9 0 243.2 109,941 109.0 Production Rate (tons/hr of unbleached air-dried pulp) 19.5 18.3 19.2 19.0 Particulates (Front-Half Catch) gr/SCF Ib/hr Ib/ton Particulates (Total Catch) gr/SCF Ib/hr Ib/ton .118 110.86 5.66 .131 122.48 6.28 .094 85.77 4.69 .107 98.25 5.37 .120 116.91 6.09 .134 131.14 6.84 .111 104.51 5.48 .124 117.29 6.16 NOTE: Standard Conditions - 70° F, 29.92 in. Hg ------- TABLE II-2 PARTICULATE EMISSION SUMMARY (Metric System) Run Number & Date 123 12/11/73 Volume of Dry Gas Sampled (Mm ) Moisture by Volume (%) Or sat Analysis (% by Volume) C00 9 °9 cS Average Stack Temperature (°C) Stack Volumetric Flow Rate (Nm /min) Percent Isokinetic Production Rate (M tons/hr of unbleached air-dried pulp) Particulates (Front-Half Catch) mg/Nm kg/hr kg/M ton Particulates (Total Catch) mg/Nm kg/hr kg/M ton 1.52 32.16 10.0 9.1 0 120.6 3,092 108.7 17.7 271.09 50.28 2.84 299.53 55.56 3.14 12/12/73 1.51 33.70 11.9 7.1 0 115.4 3,025 110.6 16.6 214.39 38.90 2.34 245.59 44.57 2.68 12/13/74 1.57 34.39 13.8 4.6 0 116.1 3,223 107.7 17.4 274.31 53.03 3.05 307.68 59.48 3.42 Average 1.53 33.42 11.9 6.9 0 117.4 3,113 109.0 17.2 253.26 47.41 2.74 284.27 53.20 3.08 NOTE: Standard Conditions - 21.1°C, 760 mm Hg ------- out from the stack. The wind cooled both the exposed length of probe and the heated cyclone/filter compartment of the sample box. This cooling was apparently sufficient to cause the condensation. The wind was moderate and relatively constant during the first two runs, but the ambient temperature increased about 20°F from run 1 to run 2. It is believed that the higher ambient temper- ature caused the reduction in cyclone condensate catch collected during the second run. The ambient temperature dropped only about 6°F between run 2 and run 3, but the wind velocity greatly increased thus accounting for the additional amount of liquid collected in the cyclone during run 3. The liquid collected in the cyclone was transferred to an individual sample container so that it could be analyzed separately. After removal of this catch from the train, the normal clean-up procedures were followed. The resulting residue weights were included in the front-half particulate cal- culations. The volume of the cyclone condensate was added to the condensate volume collected in the impingers to determine the total moisture content of the sampled gas . During the testing there was deflection of the probe when it was extended to the furthest traverse points within the stack. This was caused by the weight of approximately 10 feet of unsupported probe assembly. Prior to field testing, it was observed that when 10 feet of probe was cantilevered, the end sag was about 13 inches. The angle of deflection, 0, of the probe tip from a position normal or square to the direction of gas flow can be calculated from the following equation: - radians GEI (1) In the laboratory report contained in Appendix D, the cyclone condensate is referred to as "Front Half Water Catch." ------- where W = 0.061 Ib/in. = weight/unit length L = 120 in. = length of probe E = 2.75 x 107 Ibs/in. = Young's Modulus A I = 0.0044 in. = Moment of Inertia This angle was computed to be 8.5°. The effective nozzle area, then, is the area of an ellipse with major and minor axes of r and r times the cosine of 0, respectively, where r is the radius of the nozzle opening. As the effective nozzle area was reduced by only 1.1%, it is felt that this effect is negligible. Testing proceeded smoothly and normally on three consecutive days with no equipment, personnel or process problems encountered. ------- Ill PROCESS DESCRIPTION AND OPERATION The Alton Box Board mill at Jacksonville, Florida, produces 650 tons of kraft pulp per day. The pulp is made into linerboard in the adjoining paper mill. The EPA test program at this mill consisted of three particulate tests on the outlet of the Number 9 recovery furnace precipitator. A. Process Description 1. General Kraft pulp is produced from wood as shown in Figure III-l. In the process, wood is chipped into small pieces, then cooked in six batch digesters at elevated temperature and pressure. The cooking chemicals, called white liquor, are sodium hydroxide and sodium sulfide in water solution. The white liquor chemically dissolves lignin, leaving wood cellulose (pulp) which is filtered from the spent liquor and washed. The pulp is then made into linerboard. The balance of the process is designed to recover the cooking chemicals. Spent cooking liquor and the pulp wash water are combined for treatment. The combined stream, called weak black liquor, is con- centrated in multiple-effect evaporators and stored. As needed, the black liquor is drawn from storage and oxidized in an air sparging tank. The oxidized liquor receives its final concentration in direct contact evaporators and is then fired to a recovery furnace. Combustion of the organics in the black liquor provides most of the heat needed to generate process steam. Inorganic chemicals from the black liquor are recovered as a molten smelt at the bottom of the furnace. The smelt, consisting of sodium carbonate and sodium sulfide, is dissolved ------- FIGURE III-l THE KRAFT PULPING PROCESS o Wood Z3 CL. White Liquor (NaOH + Na2S) DIGESTER SYSTEM Piiln fe W ^ ^ PULP WASHERS eak Black Liquc Pulp Water RECOVERY FURNACE SYSTEM Heavy Black Liquor (Na Smelt Air I BLACK LIQUOR OXIDATION TANK MULTIPLE EFFECT EVAPORATOR SYSTEM 2C03 Na2S) Air 1 Water SMELT DISSOLVING TANK t T Green Liquor \ White Liquor (recycle to digester) CAUSTICIZING TANK Calcium Carbonate Mud ENGINEERING-SCIENCE, INC. ------- in water and transferred to a causticizing tank. Lime added to this tank converts sodium carbonate to sodium hydroxide, completing the regeneration of white liquor, which is then recycled to the digesters. The calcium carbonate mud that precipitates from the causticizing tank, is recycled to a kiln to regenerate lime. 2. Recovery Furnace The Number 9 recovery furnace was designed by Combustion Engineering to burn 93,750 pounds of black liquor solids per hour; this corresponds to a pulp production rate of about 600 tons per day. Fuel oil is also burned in the furnace at a design rate of about 10 gallons per minute. The furnace has a conventional design, with two parallel cascade direct contact evaporators. Strong black liquor is oxidized in a single stage, air-sparged tank. The furnace was installed in 1969. Soot is blown from the furnace boiler tubes with steam. Each soot blowing cycle takes about three hours and is performed once a shift, or less often, 3. Electrostatic Precipitator Exhaust gases from the Number 9 recovery furnace are cleaned in an electrostatic precipitator. The precipitator was installed in 1970 by the Koppers Company. The unit was designed to treat 325 F combustion gases at a rate of 243,300 ACFW. As shown in Figure III-2, the pre- cipitator has two separate chambers in parallel; each chamber has three electrical fields. The precipitator is situated near ground level and exhausts through a tall stack. Dust collecting on the precipitator electrodes is shaken loose by a system of rappers. The rappers operate in a continuous cycle, with ------- RECOVERY FURNACE AND PRECIPITATOR AT THE ALTON BOX BOARD MILL IN JACKSONVILLE, FLORIDA RECOVERY FURNACE m O z m m 3D 9 CO o m Z O m Sampling Ports Furnace Combustion Gases ^- x._ /\ nTRFPT CONTACT EVAPORATOR (EAST) ^^^ ^**\. | DIRECT CONTACT EVAPORATOR (WEST) *1 1 1 k . *-? i ELECTROSTATIC PRECIPITATOR (EAST CHAMBER) ELECTROSTATIC PRECIPITATOR (WEST CHAMBER) ^ £ INDU Air t_ Black Liquor DRAFT FAN c: 70 Z O ------- each cycle lasting 2 1/2 minutes. The dust falls into hoppers and is flushed out with black liquor, and recycled to the process. B. Process Operation 1. General The purpose of the tests was to measure particulate emission levels during normal plant operation. The information was to help demonstrate actual control levels for recovery furnace operations. As mentioned above, soot is blown from the tubes just once a shift. This practice makes it difficult to select a sampling period representa- tive of recovery furnaces in general, since most mills blow soot continually. It was decided to sample entirely during periods of soot blowing, the most severe condition for the precipitator. The -measurements, therefore, repre- sent the maximum emissions from this precipitator during normal plant operation. 2. Recovery Furnace During the particulate tests, significant furnace parameters were monitored. Readings were made every half hour and recorded on the process data sheets contained in Appendix C. As far as known from the process data and conversations with the operators, the equipment operated normally during the tests. The black liquor charge rate varied between 141 and 155 gallons per minute, as normal. Auxiliary fuel oil was fired at an average rate of about 12 gallons per minute during the first two tests. During the third, the rate was increased to 20 gallons per minute. (One of the bark boilers had broken down the previous evening, and additional steam was needed from the recovery furnace.) 11 ------- 3. Electrostatic Precipitator The operation of the precipitator was monitored during the tests. Readings of the primary voltage, primary current, and secondary current were recorded every half hour. The secondary current was read at two locations; at the precipitator control panel and in the furnace control room. Readings from the control panel were nearly twice as high, and are considered more accurate by the plant. The readings were recorded on the bottom of the furnace process data sheets, included in Appendix C. As far as known from the process data and conversations with the operators, the precipitators operated normally during the tests. Three of the discharge electrodes were inoperative, but they are just a small fraction of the several hundred electrodes installed and are not thought to affect performance significantly. 4. Equivalent Pulp Production Rate The operation of the recovery furnace is quantitatively related to the pulp production rate in the digesters. As a result, pollutant emission rates can be expressed on the basis of equivalent pulp production, as shown below: (\ / \ / /Equivalent Pulp\ Emission Rate\ _ (Emission Rate] / I Production Rate I . (Ib/ton pulp)y " y (Ib/hr) I / \ (ton/hr) / Eq. To use Equation III-l, the equivalent pulp production was calculated from the black liquor charged during the tests, as shown below: /Equivalent\ /Black \ /Pulp to Liquor\ / % Solids \ I Pulp \ = I Liquor I / Ratio \ I (Test Avg.) I Eq.(III-2) I Production I I Charged I I at avg. % Solidsl I % Solids I ' \ (tons) / \(gallons)/ \ (tons/gallon) / \(base Avg.)/ The pulp-to-liquor ratio in Equation III-2 was estimated to be 0.0022 tons of unbleached air-dried pulp per gallon of black liquor fired. The ratio was determined by dividing total pulp production for a 30-day period 12 ------- at a southern Kraft mill, by the total black liquor fired in a same period. The average percentage of solids of the black liquor was 63.2. The last term in Equation III-2 corrects for the actual percentage of solids in the black liquor charged during the tests, to that on which the pulp-to-liquor ratio is based (63.2). Substitution of the appropriate constants into Equation III-2 gives: fEquivalent\ / Black \ / \ \ Pulp \ _ / Liquor \ / 0.0022 \ / % Solids \ E , [Production/ ~l Charged I l( ton/gallon)) ( (test avg.) ) q'U x (tons) / Vgallons)/ \ / \ 63.2 / Equation III-2 was used to calculate the equivalent pulp production during each test on the recovery furnace. Dividing by the time elapsed while charging the black gave the equivalent pulp production rate. The calculations are summarized in Table III-l. As shown, the average rate was found to be 19.0 tons per hour. Substituting into Equation III-l gives the following equation, which was used to calculate mass emission rates: /Emission Rate) /Emission Rate) // 19.0 \ _, ,TTT .. ( (Ib/ton) / = \ (Ib/hr) }/ \(ton/hr))' Eq. (III-4) 13 ------- TABLE IIT-1 SUMMARY OF CALCULATIONS OP EQUIVALENT PULP PRODUCTION RATE Black Liquor Readings flours (1) Integrator (2) Avg. Z Date - 1973 Start Finish Start Finish Solids Dec. 11 1226 1600 22,624,905 22,627,954 65.5 Dec. 12 1104 1432 22,645,655 22,648,536 63.2 Dec. 13 0933 1201 22,663,017 22,665,086 65.8 Black Liquor Charged (Ibs) 30,490 28,810 20,690 Equivalent Pulp Elapsed Production (3) Time (tons) (hrs) 69.5 63.4 47.4 3.57 3.47 2.47 Equivalent Pulp Production Rate(4) (tons/hr) 19.5 18.3 19.2 19.0 (Avg.) (1) Item 12 on the process data sheets. (2) Item 11 on the process data sheets. (3) Calculated from Equation 2. (4) Calculated by dividing Equivalent Pulp Production by Elapsed Time. ------- !V SAMPLING AND ANALYTICAL PROCEDURES A. Location of Sampling Points The testing was conducted in the 185-foot high precipitator outlet stack. The work platform was about 142 feet above grade, 3 feet wide, and extended around one-third of the stack circumference. Two 4-inch pipe ports, located 90 apart, extended through the 28-inch double wall thickness and protruded 8 inches from the stack. The ports were located about 109'-6" downstream from the breeching and about 39'-3" upstream from the stack outlet to the atmosphere. The inside diameter of the stack at the test ports was 10'-9". Figures IV-I and IV-II illustrate the test site. In accordance with EPA Method 1 (Federal Register, Vol. 36, No. 247, Part II, Dec. 23, 1971), 6 point traverses were conducted because the test locations were ideally located more than 8 diameters downstream from the breeching to the stack and more than 2 diameters upstream from the top of the stack. Sampling points 1 through 6 were obtained in port A and points 7 through 12 in port B. The traverse point locations, as measured from the inside stack wall at the sampling port, were as follows; Points 1 & 7 5.75 in. 2 & 8 19.00 in. 3 & 9 38.00 in. 4 & 10 91.00 in. 5 & 11 110.00 in. 6 & 12 123.25 in. B. Sampling Procedures Because of the long probe length necessary to traverse the stack, an unsheathed, 16 ft. incolloy liner was used for the testing. It was 15 ------- GENERAL LAYOUT ALTON BOX BOARD, JACKSONVILLE, FLA, Outlet Stack Test Ports Inlet Test Ports n n n Wet Bottom Electrostatic Precip. (2) From Recovery ler Direct Contact Evaporator ------- FIGURE IV-2 ALTON BOX BOARD, JACKSONVILLE. FLA. STACK TEST PORTS Concrete Stack Ladder Port B 4" Test Ports Davit & Pulley Existing Port A Cross Section at Stack Ports Exit ID 10'6" Stack Ports ^n vo IT) f- ) _ ** 17 ENGINEERING-SCIENCE. INC. ------- fitted with a standard nozzle at one end and a standard ball-joint for connection to the sample train at the other end. The liner was wrapped with a heating tape, an insulating asbestos tape and then a protective layer of duct tape. A laboratory check of the probe heater prior to the field work indicated the air inside the tube was heated to 270°F when plugged directly into a 110-volt AC supply. This same power source was used during the testing program. A Type S pitot tube, approximately 15 feet long, was clamped to the probe completing the assembly. An overhead angle rail was set up from which the sample box/probe assembly was supported by a trolley. This allowed for smooth in and out movement of the probe during the long traverse. A standard EPA sampling train including a cyclone separator was utilized during all testing. The cyclone was installed upstream of the filter holder, inside the heated box. Two passes were made in each port so that a total of 10 minutes per traverse point and 2 hours per test total sampling time were accumulated. The sampling was conducted as specified by EPA Methods 1 through 5 with the additional requirement that the cyclone condensate and the impinger contents were collected and analyzed for particulate content. These methods were published in the Federal Register, Volume 36, Number 247, Part II, Thursday, December 23, 1971. The procedure for recovery and analysis of the impinger contents are published in the Federal Register, Volume 36, Number 159, Part II, Tuesday, August 17, 1971. The cyclone condensate was recovered in the same manner as the impinger contents. C, Analytical Procedures The cleanup site was located in an enclosed truck which was parked yery close to the base of the stack. All sample train preparation and 18 ------- clean-up was performed in this clean, well-lighted, enclosed area. The particulate trains were charged, capped and hoisted up to the stack work platform by rope and pulley. After each test, the train and probe were capped and lowered to the ground. They were then carried the short dis- tance to the truck and cleaning proceeded. All the integrated gas samples, taken during particulate sampling, were run through the Orsat analysis the same day they were collected, soon after the test run was completed. The procedures followed for the laboratory analyses and calculations of results are defined in the previously referenced issues of the Federal Register. 19 ------- |