Document k64m5aBNQK9ay3o2oRvpNEX1O

S\ *' f' ***** R. A. Conrad Aberdeen Plant PVC Yield May 22, 1978 Page 2 A4**.* '* CONFIDENTIAL SUBMITTED UNDER AmftNEY-tnTFT PRIVILEGE 4. Since mid-February, the plant has been operating the equipment associated with the Capacity Replacement Project. Substantial losses associated with operational problems have occurred in this unit. During April, failures occurred due to a) Reactor rupture disk-relief valve, b) Recovery system sight glass, ____________ c) Reactor-condenser gasket, and I | .( { j AI d) Reactor agitator blades i,* tUL.i i I f The April yield study involved the following work: " 1. Received vinyl chloride railcars were each weighed into and out of the plant to verify VCM input. 2. Daily VCM inventories were calculated and compared to batch charge amounts. I* 3. Three dryer stack emissions tests were conducted based upon EPA approved methods. * 4. Water samples were obtained for PVC solids content: a) Dryer centrate samples - 25 per month b) Reactor rinse samples - 15 per month c) Blowdown tank drain samples - 40 per month d) Total process effluent samples - 32 during month e) Samples of Chem Wash - 2 during month 5. Inventory verifications were made daily of dry blend and compound resin usage. 6. Accounting calculation methods weTe checked to insure accuracy. The final report containing detailed 4nalyses and calculations of the April work is in preparation. A copy of the full report will be sent to you when complete. The essential results are indicated below. Through process sampling and monitoring, the majority of process losses were identified. Approximately 89% of the 986,381 pounds can be located. These losses are listed on the attached table. Items 1 to 5 (156,000 pounds) were caused by equipment or instrument failures. These are essentially non-recurring losses. Items 7 and 8 (17,200 pounds) will require mechanical improvements to eliminate. Action steps are in progress or have been completed. SAL 000089953 iXtf edftfi'api R. A. Conrad Aberdeen Plant PVC Yield May 22, 1978 Page 3 CONFIDENTIAL SUBMITTED UNDER ATTORNEY-CLIENT PRIVILEGE W: * * - r Item 9, reactor rinse loss (116,800 pounds) is, to an extent, inherent in the process. This can be minimized by operator attention. Operations now samples and grades operators based upon solids content in rinse water. **ri S * j v -li jpi\| j j j\\ -: * 3` * . ;.. S`1 -i .?F--/' Item 10, foaming loss, is the largest loss identified (268,000 pounds). This loss was caused by the failure of both recovery inlet control systems. Additionally, the attempt to maintain a S.5 minute dry time also amplifies foaming. The control system repairs are complete. OMS addition could negate the foaming tendency at low dry times. At the level determined Items 11 and 12 (33,300) are inherent in the process. Items 13 and 14 (90,000) are also inherent in the process. These dust collectors represent environmental design treatment technology. Items IS and 16, (15,300) are acceptable and are the slurry and water residual VCM content. Item 17 (80,100) was excessive due to a refrigeration system fail ure. Tie-ins have now been completed and this loss is now less than 6000 ppunds per month. At this level, this is an inherent process loss. Items 18, 20, 22 and 23 (7500) are inherent losses for which no corrective actions are feasible. Item 19, (46,500) is somewhat high. "Wearing In" of reactors seems to occur after startup. During this period cull formation seems to be high compared to normal operations. This was observed with D-500 and D-600 as well as at Oklahoma City. We are now experiencing this - with the CRP reactors. f r. Item 21 (41,600) is an estimate based upon the loss of one pound per bag of material. This estimate seems realistic from visual observation of the losses encountered during debagging. In summary, the study during April did not identify any major, single, cause of yield loss. The sum of all identified losses did account for about 90% of the material lost. Feasible corrective actions have been taken. The greatest identified loss, foaming, will require a pro cess operational change to correct. This change (OMS addition) will be a 1979 Capital Budget project. A means of selective OMS addition is being installed and when operable will reduce the nature of this loss. We are' continuing to sample the mentioned loss points. If any major losses occur corrective operating actions will be instituted. SAL 000089954 R. A. Conrad Aberdeen Plant PVC Yield Nay 22, 1978 Page 4 CONFIDENTIAL SUBMITTED UNDER ATTORNEY-CLIENT PRIVILEGE We have also requested assistance of PED and R 8 D on this problem. We have asked for manpower for a week long process surveillance test. This would be to verify that all major loss sources have been accounted for. As of now we expect this surveillance test to be conducted the second week in June. Douglas Michels Chief Process Engineer :tap ` c: \^D- ELK-DWF-JFH \\ K..J \: * > / SAL Q00089955 . ,*.^1 * v ?***e^4*0*4 4 TABLE I APRIL PROCESS YIELD LOSSES ABERDEEN PVC PLANT \VA> ** ' CONFIDENTIAL SUBMITTED UNDER ATTORNEY-CLIENT Privilege- 1. 2. 3. 4. 5. 6. 7. - 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 530S Coarse Batch Due to Agitator Failure (4) 5425 Coarse Batch in D-400 (4) Reactor Rupture Disk-Relief Valve Failure (4) Reactor-Condenser Gasket Failure (4) Recovery Line Sight Glass Blowout (4) Blend Tank Overflow (4) 407 Blend Tank Agitators Seal Failure (4) CRP Dump System Control and Leakage Problem (4) PVC Loss in Reactor Rinses (1) Losses Due to Batch Foaming (1) Losses of PVC In Chem Wash (1) PVC Lost in Dryer Centrate (1) Dryer Dust Collector Losses (1) SiloDust Collector Losses (3) VCM Loss Due to Slurry Residual Content (1) VCM Loss Due to Blowdown Water Residual Content(1) ' VCM Loss Due to Receiver Venting (3) VCM Loss Due to Reactor Opening Loss (1) * Reactor Culls (2) PVC Dusting During Bagging (4) PVC Lost Due to Debagging (4) PVC Lost During Railcar Loading (4) PVC Consumed in Process Sampling (2) TOTAL (1) Based upon the average of all samples obtained during April. (2) Based upon weighed amounts. ^ (3) Based upon equipment design data. (4) Based upon observation estimates. 36,000 25,000 25,000 20,000 30,000 20,000 2,000 15,200 116,800 268,000 14,500 18,800 70,500 20,000 10,300 5,000 80,100 2,000 46,500 500 41,600 5,000 4,000 876,800 S*L 00089?56 i * 7- ij/ f I. Propane Vaporizer Replacement fere** CONFIDENTIAL Submitted under ATTORNEY-CLIENT PRIVILEGE The completion of the process design awaits final vendor data. J. Plasticiser-Stabilizer Weighup System Pf p'*r'r0.rfj ( Revisions to this design are in progress. K. Plasticizer Chemical Exposure Abatement * The plant reviewed the process design by PED. CED is now preparing a definitive estimate. L. Plant Yield Loss Study An extensive effort was made during April to locate and define all sources of VCM or PVC yield loss. Some of the steps taken were: 1. Each railcar of VCM was weighed prior to aftd after unloading. 2. Vinyl chloride inventories were obtained daily and. compared to the tabulated reactor charge sheets. 3. An attempt was made to quantify the four major non-recurring losses during the month: a) 744 Coarse 5305 batch --45,000 pounds b) Recovery System Sight Glass Blowout - -15,000 pounds c) Reactor Relief Valve Emission * - 30,000 pounds d) Reactor Flange Blowout - 30,000 pounds 4. Periodic samples were obtained of chem wash and rinse water to look for resin losses. 5. Periodic samples were obtained of blowdown tank water to look for; resin losses. * 1 ' r * '1 * 6. Periodic samples were obtained of secondary influent water to look for losses. 7. Estimates were made of spills from the dump systems and blend tanks. 8. Dryer stack emission tests were conducted. 9. Centrate samples from each dryer were obtained weekly. 10.- Dry blend usage is being determined several ways. 11. Each railcar and truck of PVC isbeing weighed out. All of-the-*data taken during April is now being analyzed, A report will * be published late in May. 0oOOB*