Oklahoma department of environmental quality

**Two-year average ADT

Emissions of VOCs from SC-1, solvent cleaning, are based on the use of a 100% VOC solvent to clean Paper Machine wires. This solvent is applied through spray nozzles located across a boom that stretches across the Paper Machine. Emission amounts authorized in the pending Part 70 permit for machines PM-11 through PM-14 are based on research from a similar de-ink facility paper machine. According to that research, only 60% of the solvent applied to the machine becomes air emissions, 20% is consumed in the reaction with the latex buildup on the wire, and 20% ends up in the water loop and eventually in the wastewater treatment plant. This 60% evaporation rate analysis was first utilized to calculate authorized emissions in the memorandum associated with Permit No. 91-127-O (M-1), issued for PM-15. Extending this analysis to the other four machines that had not previously been subject to emission limits would have resulted in authorizing 179.1 TPY of VOC for a rolling 12-month total, but the facility requested a limit of 338 TPY, based on historic increases in use of such solvents. Permitted emissions of 37.57 TPY of VOC for PM-15 were carried into the pending Part 70 permit from Permit No. 91-127-O (M-1).

In estimating emissions from the Poly Plant extruder processes, emission factors from a 1996 article in the Journal of the Air and Waste Management Association (JAWMA) were used for linear low density polyethylene (LLDPE) blown film at a 355F melting temperature. The polyethylene (poly) film produced at the Poly Plant is a blown film process using 20 – 22% LLDPE at a melting temperature of approximately 350F. Other emission factors associated with the JAWMA document refer to either high-density blow molding or extrusion coating. Neither of these processes matches the process at the Poly Plant.

These equipment items generate a corona that is used to treat the polyethylene film produced at the Polyethylene Plant, allowing for enhanced ink bonding to the film. Ozone is generated by this process and is vented to the atmosphere without any controls. Manufacturer’s data for the existing units indicates design production of 0.073 lbs/hr/kVa, where kVa is kilovolt-ampere. The principal unit is rated at 5 kVa. A backup unit is rated at 10 kVa and has an associated decomposer, which converts 95% of the ozone to molecular oxygen. The Part 70 permit memorandum shows total ozone emissions from these two treaters to be 1.76 TPY, assuming continuous operation. Three new units to be added in the current project do not have decomposers, so their emissions are simply

3 units  10 kVa/unit  0.073 lb/kVa/hr  8,760 hrs/yr = 9.6 TPY of ozone.

This entire amount may be treated as the increase in emissions for this equipment.

Analysis in the memorandum associated with the pending Part 70 permit estimated VOC emissions for the process to be 1.83 TPY for a single press, based on solvent component concentrations and the fact that the solvents are 100% VOC. Emissions are small because almost all of the used solvent is recycled at the Poly Plant. Further review of actual use suggests that 900 lbs of VOC per year is a more reasonable number. Adding three more presses will not multiply the emissions by four, because the plates made may be used interchangeably among the presses. Despite that fact, the facility has opted to make the conservatively high assumption that emissions will be 900  4 = 3,600 lbs/yr (1.8 TPY). This number is less than the amount estimated in the Part 70 memorandum, so the actual-to-potential difference is taken to be 900 lb/press/yr times three new presses, or 1.35 TPY.

VOC emissions from polyethylene printing depend on capture and destruction efficiency. The existing single printer has capture efficiency of only 70%, which is a datum agreed upon by the facility and DEQ in discussions surrounding the Part 70 permit. Recall that testing indicates 96.8% destruction efficiency. Average 2002-2003 uncontrolled VOC from the printer was 197.5 TPY. Fugitive emissions were 30% of this number, or 59.3 TPY. The 70% portion captured was reduced by 96.8%, leaving stack emissions of 1.9 TPY. Plans for the new printers include constructing an enclosure around all four units, leading to 100% capture. Assuming conservatively low destruction efficiency of 95% and using potential printer VOC emissions of 971 TPY leads to zero fugitive emissions and 48.6 TPY of stack emissions. The following table summarized these calculations.

VOC emissions from paper printing are calculated based on assumptions made in analyses for earlier permits and assuming that the current project will lead to a large increase in the amount of printed paper products manufactured. The project will increase actual paper production, but will not exceed the amount of paper already authorized. Thus, the existing 92.28 TPY of VOC authorized by existing permits is not to be increased by this permit. At the time the VOC level was authorized, average VOC concentration in water-based inks ranged from approximately 6% to 8%. A conservatively high 10% was used to calculate a PTE and to allow for flexibility in varying ink VOC concentrations. The data used imply use of 0.343 pounds of VOC per ton of paper. Actual use figures for 2002 and 2003 indicate that calculations supporting the 92 TPY level greatly exaggerated the VOC concentration, in that 2002 use was approximately 0.018 lbs/ton and 2003 use was approximately 0.031 lbs/ton. There are two possible approaches to establish the actual-to-potential increase. First (Option 1), we could use the original emission factor of 0.343 lbs/ton and apply it to all potential production, comparing that result with 2002-2003 actuals. Second, (Option 2) we could apply the 0.343 lbs/ton emission factor to only the increase in production, holding the existing production at the levels shown in 2002-2003, and then compare the total to the 2002-2003 actuals. Both options use actual average production of 345,880 tons and design capacity of 538,845 tons per year, for an increase of 192,965 tons per year. The following table shows these calculations. The second option is clearly more realistic, but the first option may be required for technically accurate PSD analysis.