Document 4aXp3Jk0O7KdB1nRmbz2KNqwp

1o Code Zero Steering Committee FIELD POINT 0.1 AK.UVI .JIPAHTMCNT tk ULC'O. NO. Cleveland ~ ^ 6*AT YOJ.I LUT i FROM F. E. Krause AUDJCCT [FEK-194-74] FIELD rOIN r OR AKRCrt nri'ARTMC'.MT Pi m.r >G. MO. . --. Avon Lake Technical Center t y , -- sDf. rlT'THI'i LI. TT:t November 26,1974 ^ v v4 ,\ Code Zero Task Force - Bi-Weekly Progress Report* s~-32rC:N S CLOSED POLY GP/EP Clean Poly Recipe. We had a successful charge in the ALTC 1100-gallon reactor at 1.4/1 water to monomer ratio (Paper 0, Sand 0.5 - .lightly scattered in upper section). Granulation and porosity was good. Primary changewas minimum delaynliv charging. We will be discussing a plant trial with AL3C next wfe^kvin 1100 or B^O-g^iipjK reactor. Primary production concern will be hydroful^&eratiow. WitenhaMr\eports^oO ppb Code 11 in the resin with 5 ppm charged in the r^cqp\. This should da^^rtely cause no concern in toxicity. v Closed Poly Demonstration. The latest in Geon-East was terminated after lSf^barges to inst for the agitator hub and a reinforced cement bo The reactor will be recoatei&ajad a simulated Cli gallon\jlass lined reactor in shecrstainless steel cover easy bottom drainage, initiated. Backderf-Pancallo-Summers) The Louisville Closecki&iy run now stands a1 arges and is, beyond all doubt, a success. By coating wi^n/Backderf's d Bondex cement we have solved the trouble spot budUOTu problem. DeyelcJ now turn control of the remainder of the run back to ttraSplant. The ned for inspection after Charges 1, 12, and 30. The poly had to be open^ ChaisgS 21 to clean out clogged spray nozzles. The poly has not been entered, j.n any way except for cleaning the nozzles. The only buildup in the polyis iple Mf lumps about 2" in diameter on the bottom steady bearing supports and so^\pap e buildup on the bottom backside of the baffles. The rest is clean! (Witenhafer) Coatings. The 16-(gallon"Tractor coated with a nylon-Code 11-zinc oxide coating has shown no buildup\for\5V10 charges in the top of the reactor (minimal agitation) and no buildup fof^tKchaJR^rajon sides of reactor (maximum agitation). Code 11 analysis on several charjwa^shows: 1st Charge 2nd Charge 3rd Charge 9th Charge 10,400 ppb 2,500 ppb 1,240 ppb 150 ppb Various methods of extracting Code 11 from a fresh coating are being investigated to determine if the initial high level of Code 11 can be reduced. The water insoluble Code 11 will also be investigated. (Antlfinr.er) Geon 92 reactors coated with latex modified cement continue to show promise. Poly l4l run was terminated after 45 charges when the reactor was accidentally aquablasted. The 24654001 UKC>-4i:.<S-S HF.V. U 70 LiTHO. IN U.fr.A. BFG33321 Page 2 - Code Zero Task Force Bi-Weekly Progress Report - 32 November 26, 1974 reactor had been cleaned with the HRC (no aquablasting) after each charge protecting the cement bottom with a blanket of H^O. Four poly entries were required to clean buildup out of the overhead nozzles. After cement was removed by aquablasting, it is again necessary to aquablast every 1-2 charges. The floor coating will be replaced. In addition, the overhead nozzles will be coated with the latex-modified cement. We hope to demonstrate that we can go 50 to 100 G-92 charges with no aquablasting and no poly entries. (Costantino-Backdcrf) G-92 Ebly l4o which was also repaired with cement/latex coatings has been through 23 charges. Each was HRC'd with water in the bottom and then coated with Code 1QA. The top half of the reactor (water contact) shows no buildup on the cement. There is no damage to the cement by HRC. This excellent adhesion was achieved by sand blasting the chipped glass areas, then painting on a coating of Bondex cement/latex. The walls on the bottom half (vinyl chloride contact) show a loosely adhered buildup which occasionally comes off and causes a plugging problem in the drain. This cement is soft and crumbling and has been removed in some areas. The bottom, which is a heavier cement/latex coating, is showing some loss of adhesion and some softening. Cement has come off in several small spots. This problem may relate to higher latex in the cement which swells with VCM and starts to soften. Build up on the cement appears not to grow. Cement in the thermowell is standing up well. The buildup that has formed appears not to grow. Cement on the agitator hub is holding up similar to that in the thermowell. (Costantino-Backderf-Summers) For problem buildup areas, sandblasting - coating with a latex modified cement plus a top coat of Geon 44-3/Code 1QA appears to be an excellent solution in GP/EP reactors. Evidence is 18 charges in Poly l6, Louisville, 12 charges of test panels in 103EP F-76, and 23 charges in the top half of G-92 poly. (Summers) Experiments by D. E. Witenhafer have shown that amino groups increase solubility in VCM, whereas hydroxy functions depress solubility. Further, amino groups appear to have greater effect on preventing FVC buildup. On this basis the reduction of 3>3'- dinitrodiphenylamine to 3,3'-diaminodiphenylamine was undertaken. The reduction, by hydrogen over a Ni catalyst, was unsuccessful. AH that I obtained was a little tarry material and much cinder-like residue. The synthesis will be repeated in the absence of air. (Kehe) ZQ O V^^z The concept of water soluble inhibitor coatings is attractive from an application view point. Such a coating might be sprayed on with hot water and show greatly reduced solubility at the polymerization temperatures, and might also be VCM repellent. A low temperature (150C) condensation of gluconic acid (a penta hydroxy acid) and m-PDA gave a light brown tough resin. It is soluble in water, partially soluble in methanol, insoluble in benzene and chloroform. Applied to FVC poly surfaces the coating may cure upon heating and become water insoluble. The OH groups on the chain could then be VCM repellent due to their OH groups. The latter preparation will be tested in the near future. (Kehe) Preventive Cleaning. Our first series was terminated after ten cycles to clean the dome area and some trouble spot areas which showed some buildup. The intent is to determine whether buildup in these areas are due to inadequate cleaning, surface rough ness or stagnation in agitation - non-hydroful. Antifinger's nylon/Code 11 coating will also be tried at the liquid level surface. Cleaning these areas.was a. difficult 2-day job. We have now started the second series but bow-----into some problems. An BFG33322 Page 3 - Code Zero Task Force Bi-Woekly Progress Report - 32 November 26, 197^ oil seal leak appears to have contaminated our cleaning solution and we are presently trying to reformulate the solution so it again provides water wettuble stainless steel. A possible alternate to this problem in the future is the use of a water miscible glycerine based lubricant instead of oil in the seals. This may also minimize some problems in polymerization like no initiation seen in the past with oil leaks. This will be pursued with Engineering. Resin properties in our first series were normal at standard water/monomer ratios indicating we need not be concerned with colloidal problems encountered in 3-liter work. Also, the change in the surface of reactor glass panels aged at 90C in 10$ Alconox for > 88 hours (opalescent coating) was surprisingly nearly the sane using D.M. water. On the other hand, the TSP solution currently being used definitely attacked the glass producing deep pits in addition to the opalescent effects which are minor in comparison to pitting. I would say we should develop acidic cleaning solutions for use in prevent ing buildup on chrome super baffles and stainless steel agitators in glass reactors. Until then Alconox could be tried for a short term evaluation of the principle.(Cohen) Dispersion Resin - Geon 121 Type A large full scale production run of Geon 120X283 has successfully been completed at Henry. A dryer study has also been made at the plant to optimize drier-grinder condi tions for this new resin primarily for lowering plastisol Brookfield viscosity age. The drier study data is currently being analyzed. The run consisted of 60 charges with 10 charges per poly using only HRC cleaning between charges. The polymerizations went smoothly with excellent reaction controls. A temporary problem with a low pH soap solution was quickly corrected. Low poly buildup and the absence of "chunky" coagulum were consistent throughout the run. Drain line plugging, the major imminent Code Zero problem with Geon 121, was completely eliminated. After 10 charges the polys began to show signs of buildup -' about what is typical after one or two charges of Geon 121. (Mikofalvy) Geon 128 Type. A substantial advance has been made at Henry in bring the Geon 128 type polymerization process tinder control. The work was done on Geon 12CK271, our high gloss, heat stable, blush resistant flooring resin. Several procedural changes aimed at controlling pH, surface tension and vacuum have been implemented giving good and consistent initiation, more reproducible reaction rates and good temperature control at 35$ total solids polymerization. As a result, thick charges have been eliminated, coagulum level cut in half and the polymerization cycle time reduced to 22 hours. Thes^ are encouraging results compared to previous runs in July and August at 40% TS polymer-Q^ ization with 36 hours cycle time, high coagulum level and frequent thick charges. Withth additional plant process work it seems quite feasible to attain a further reduction in`k coagulum level and cycle time. (Mikofalvy-Pyle)g Latex stability during the polymerization of Geon 128 is affected by soap coverage and^ by pH, which controls the effectiveness of the soap. Control of .soap coverage can be BFG33323 Page 4 - Code Zero Task Force Bi-Weekly Progress Report - 32 November 26, 1974 upset by the initiation of new, small particles. New-particle initiation is much more likely if the soap coverage is high (80%> or more) and if the number of seed . particles is small. Thus the worst case is the growth of 120X198 seed to 120X199, the large component of Geon 128. We have helped this situation by using a mixed seed, containing large seed particles for the large component and small seed particles for the small component. The number of small seed particles in this mixture is very large, and this does suppress the new particles. Using this mixed seed we find we can get away with as little as 6($ soap coverage, though 70%> is more comfortable. (Paxton) We have been stressing the importance of "high enough" pH for Geon 128 latex stability. But what pH value is "high enough"? From base titrations of partially acidified K la-urate solution, I find that a pH of over 9-5 is needed to make the soap solutions clear. This is true using either KOH or NH4OH. (Paxton) Latex. Production of improved quality Geon latex at AIGC has reduced the number of strainer changes, thus reducing VCM exposure. Continued studies in development are aimed at optimization of the seed latex recipe and at elimination of prepolymerization in monomer premixes. (Barden-Shah-Weaver) Parts are on order for improved ventilation and vessel purging systems at ALGC. Mean while, a temporary vent system is in operation at two strainers in the stripping system. The vent system is being expanded to other strainers. (Barden) POST POLYMERIZATION gp/ep The stripping schedule programmed October 17, 1974 has been completed. Scanning electron microscope photographs have indicated no significant visual difference when comparing particles from normally processed slurries with particles from slurry flashed into the stripping column at temperatures 50F above glass transition temperature. Screen analysis of resin from isolated raw slurry, stripped product indicates narrower particle size range and movement toward a coarser average particle size than that of normally recovered, recent production of the same product type. We have isolated close to 6,000 pounds of nil RVCM raw slurry column stripped 11CK400 product for customer evaluation. RVCM desorption work by Chandler at AITC indicates column stripped 110X401 product as "very similar" to normally produced 110X401. The stripping column feed cyclone was used for just a few hours in the program. Conditions of operation and product types stripped precluded the need for a feed cyclone. Anticipated near future processing of 11CK334 slurries may be a performance test of the cyclone - although Cleveland Engineering claims diameter should be equivalent to column diameter for adequate disengagment. We are presently trying to define what additional work should be done to develop operating experience and optimum design of this column.(Grellinger) The E.A. has been approved for installing a total of seven slurry stripping columns at Long Beach, Henry, ALGC, and Pedricktown and Girdler has been advised to proceed with procuring equipment and materials. Final details have been worked out with Girdler concerning the package skid arrangement. Work is proceeding with the plants on defin ing interfacing requirements and drawing up flowsheets. Particular emphasis is being given to establishing requirements for the recovery equipment. (Calsing) BFG33324 Page 5 - Code Zero Task Force Bi-Weekly Progress Report - 32 November 2o, 1974 Dispersion and Blending Resins Some plant scale stripping studies are being carried out to determine what levels of RVCM can be obtained with present equipment. Equipment modifications and techniques are being studied to improve the removal capability. (Huddleston-Linak) The Henry plant does all recovery in the polymerizer. Normally the slurry is heated to about 170F using jacket heat with vacuum on the vessel. This yields slurry with 2000-5000 ppm RVCM. In some tests with Geon 128, temperatures as high as 210F have been obtained with RVCM levels below 400 ppm. There was some thickening and discoloru tion so they feel that temperature should be limited to about 190F. (Edwards-Pyle) Avon Lake plant stripping is a one-stage flash from the blowdown vessels into an evacuated stripper. Part of the monomer is taken off in the blowdown by reducing pressure to about 1 psig. In a series of tests the effect of different temperatures in the line to the stripper on monomer removal has been established. With Geon 121 the level can be reduced to 250-500 ppm in one flash by heating to about 210F. The tempera ture-monomer relationships fit very closely those established in overall scale tests at ITC. Methods to achieve a second flash are being considered. (Laughlin) Louisville has a two stage stripping system. In normal operation with l40F on the first stage and no additional heat to the second stage the product shows about 5000 ppm. Minor modifications are being made to evaluate higher temperatures. (Russell) Pedricktown does all recovery in the polymerizer with results similar to those obtained at Henry. The steps necessary to activate their stripping system are being considered. Some very interesting results were obtained in a modified Pieco stripper with Geon 121. This stripper is essentially a two-stage flash stripper with provision for mixing steam with the slurry before it passes into the separator or flash tank. Originally there were plate heat exchangers before each stage. These have been replaced by pipe with two restrictions in series to improve a back pressure and condense most of the steam. From the restrictions the stream expands with separation of vapor and liquid to some extent followed by complete separation in the flash tank. At l60"F in each stage RVCM was reduced from 9157 to 1314, at 175 F to 619, at l89F to l4l and at 200 F to 86 ppm. These values for two stages fit very well with the data from small scale tests at ITC. These tests were made on drum quantities since the stripper is located in another building away from the polymerization equipment. A more extensive test is in progress with a tank truck of slurry. This test will involve a range of temperatures, flow rates, vacuums and enough product will be segregated to dry and evaluate. (Doughty) The wetted-wall column stripping of Geons 121 and 120X283 looks promising in prelimin- are experiments but have no firm data as yet. Early qualitative results indicate a four decade reduction in the area of the VCM peak of head space samples taken over slurry samples. Exact removal results need to be obtained. Foaming is still a problem, even in this low shear device. However, the foaming can be controlled by apparatus design and feed rate control. (Davis) Column stripping of Geon 202 and 101F-4 blending resins is completed. The results with 202, as reported previously, were good. Analytical results just in on Geon 101 F-4 are also very good. We have reduced the RVCM in 101 F-4 to trace levels (< 10 ppm) in ~ 10 minutes residence time. The 101 F-4 handles very nicely in the column. I would have to say that column steam stripping of 202 and 101 F-4 is feasible and effective. Recycle stripping of Geon 202 will be evaluated as soon as I receive a resupply of Geon 202 wet cake. (Davis) SOOVQ BFG33325 Page 6 - Code Zero Tack Force Bi-Weekly Progress Report - 32 November 26, 1974 Latex ALGC is stripping Geon 352 and Geon 460X6 latex by a recycle flash operation to less than 100 ppm RVCM. There is some dilution and considerable loss of productivity. Lab work indicates that effective chemical stripping of VDCM latexes is practical. A redox tail experiment reduced the RVDCM to nil. A charge bottle has been installed on one poly to test the method in production. (Barden) Two more sensitive latexes produced at ALGC have been multi-stage flash stripped successfully using steam injection in 50-gallon equipment at ITC: Geon 353X3 seed and Geon 460X1, a self curing latex. RVCM levels after four passes were around 10 ppm. Other tests have to be made prior to recommending operating conditions for stripping in the plant. (Weaver) Geon 35CX2 is our most sensitive latex. All attempts to heat by direct steam injec tion, however slow, have resulted in gross coagulum formation. Louisville has stripped an 8000-gallon blend to 500 ppm RVCM by recycling for .32 hours through a Rodney-Hunt evaporator at a jacket temperature of 170F. A higher temperature will be tried on the next run. A batch of Geon 352 was stripped to 55 ppm RVCM to convert to Geon 580X52 for shipment to American Cyanamid. We hope to get AVCM measurements during a wall covering fabrica tion run in their plant. Carbon tube measurement at the drumming station in our plant showed AVCM levels no higher than the ground level of 7*8 ppm. Geon latex is preplasticized by agitating a mixture of resin latex and plasticizer for a specified time at high temperature. Batches are normally cooled by running cold water through the jacket. We are attempting to take advantage of the high temperature to remove RVCM by flash cooling under vacuum. The first experiment showed that one- half the RVCM was removed. With better control of the vacuum and of the latex flow we should be able to do much better. (McAlister) At Louisville arrangements are proceeding for insulating the stripping vessels. The estimate for the ALGC stripping job was completed and is now being reviewed by plant and engineering personnel in an effort to reduce the cost of this work. (Calsing) Hi Temp A plant trial at Louisville proved the feasibility of removing IlCClg and CCI4 from Hi Temp resin by batch stripping for one hour at l40C. The levels for HCCl^and CCI4 went from 3.87% and 2.8l$, respectively, in the starting slurry to 0 and trace (ca 200 ppm or less), respectively, in the stripped resin. The stripped resin was discolored but heat stability was equivalent to the control. (Boyer-Ballman) We have extended the pressure stripping work on Hi Temp to still higher pressures (50-60 psig). Although we obtained lower residual CCI4 levels, the results are still off the mark. Typically, at 60 psig we obtain a decrease in CCI4 from 12,000 ppm to ~ 1,000 ppm with a 20 minute residence time in the 6-inch column. This is poorer than anticipate from lower pressure results obtained earlier. Verification of the results, extension to higher pressure, and an evaluation of the effects of pressure stripping on resin stabilit will be made as soon as we can resupply our Hi Temp. (Davis) Several alternate neutralizing compounds, to soda ash, have been evaluated to find one that effectively reduces discoloration of Hi Temp during stripping. CuCOg, alumina and 9 Q O t 9 $ Z BFG33326 Page 7 - Code Zero Task Force Bi-Weekly Progress Report - 32 November 26, 197^ NaHSOo all are effective in reducing discoloration and Hi Temp samples neutralized with these compounds look equivalent to the control in heat stability tests. CaO and MgO produced some discoloration during stripping but significantly superior heat stabil ity as compared to the control. Other compounds are being evaluated. (Boyer) Post Treatment Work has been centered around closing the material balance on the chlorine in the KMnO^ oxidation of VCM. The chemistry of the oxidation has been firmly established, but quantitative results have been pointing out less chlorine accounted for than predicted from Berens' solubility data. (Daniels-O'Mara-Dorsch) HI CONVERSION UPS-1 Plant Status. All Plants continuing to work toward conversion of Geon 102EP F-5, 103EP, and 103EP F-76 to UPS-1 by the end of '75. No significant known polymerization problems. Electricals at ALSO appear to be down while Long Beach is equal to the control. Addi tional testing on subsequent lots will be necessary before we can make any definite conclusions. (Marty) R&D. Work is continuing with each of the Plants to zero in their UPS-1 recipes. We have not made an M.A. run yet on the improved bulk density version at ALGC-East.(Yieh) UPS-2 Prognosis on a Geon 90 replacement is still not overly optimistic. Although 30-gallon results are more reproducible we still see uncontrolled variation in particle size and PSD. Reducing particle size to ~ 200 microns improved uniformity but reduced porosity to 0.30-0.33 which we believe is too lew to compete with G-90. NKEP tests will be run to confirm this. More attention to actual charging variation with some planned varia tion may help tie down problems in control of initiation and agglomeration. (Kemp) Three-Component Dispersant Systems There are seven 3-component dispersant systems being screened. Four of the seven are completed, and two of those are judged to be inadequate. Those two are Vinol 5^0 + Methocel + Tween and Code 7 + Methocel + CaLS. The Code 7 + Vinol 5^0 + Tween and the Code 7 + Methocel + Tween combinations deserve a little more work. There are, of course, three systems yet to .go in the original 35-charge program. (Morgan) END PRODUCT TESTING The public warehouse monitoring program is underway. Samples have been taken at Kelmer warehouse (Chicago, Illinois) and anlyzed by GC at ALTC. A total of twelve samples taken around storage areas for dispersion resins (Geons 121 and 128 in bags) and latex (Geons 35^, 580X12, 576 in fiber and steel drums) showed that the highest levels measured were 0.6 and 0.7 parts per million (ppm) AVCM several feet away from fiber drums containing Geon 576 latex. All other samples showed considerably less than 0.5 BFG33327 Page 8 - Code Zero Task Force Bi-Weekly Progress Report - 32 November 26, 1971* ppm. Several more warehouses and a bulk resin transfer facility will be monitored. (Kruszynski-Solak-Dorsch) The Distribution Department has uncovered a need to collect AVCM data relative to less- than-truckload (LTL) shipments of PVC products. This data is necessary to determine what levels of exposure, if any, are present to truckers or warehousemen involved with LTL shipments and, hopefully, to permit these operations to be excluded or given some variances under the OSHA standard for VCM. An experiment was conducted at AIGCP in which three previously empty trailers were loaded with one skid (2500 lbs.), two skids (5000 lbs.) and four skids (10,000 lbs.) of Geon 92 resin which had an RVCM level of < 10 ppm. The trailers were closed for 72 hours, re-opened, and AVCM samples collected. GC analysis of these samples showed that the highest level of AVCM present under these conditions was less than 0.3 ppm. Additional experiments will be done with rigid and flexible compound. (Rider-Begnaud-Solak-Dorsch) ANALYTICAL A procedure for using the Perkin-Elmer Headspace Analyzer for personal monitoring samples is now being finalized. Our current plans are to put this method into practice at AIGCP the week of November 25. Training of ALTC personnel in using this procedure is also now underway and it is anticipated that we will start using this method at ALTC the week of December 2. (Whitney-DeCapita) A Bendix Flasher System (including a gas chromatograph with backflush valve) has been evaluated at the Bendix plant and is now being further tested in the Manufacturing Services Laboratory. Both of these evaluations look very encouraging with respect to providing an addition, rapid method of analyzing personal monitoring samples. (Whitney-Salzer) An accelerated program is now underway to develop an automated headspace method for the analysis of RVCM in latex. The initial results from this program also look very good. Our objective is to complete this new procedure within the next two weeks. (Mele-DeCapita-Whitney) A new testing program to determine the efficiency of carbon respirators and cartridges is now in progress. Our objective is to screen all of the various types (and different suppliers) of respirator devices now in use for VCM. This testing is now being done in our Physical Test Laboratory and is being established as a routine test. This total testing program will require about three months to complete. (Baron-O'Mara) r FEKicl F. E. Krause Distribution List Attached BFG33328 80Qt7S9frz CODE ZERO BI-WEEKLY PROGRESS REPORT DISTRIBUTION CLEVELAND W. F. Bixby E. B. Osborne * W. E . Brodine H. R. Rex H. R. Calsing R. J. Coffey M. E, Roha G. D.Schaaf-E .R.Clayson * B. A. DiLiddo * R. D. Scott C. R. Flynn * E. G. Scbwaegerle L. B. Gajdos E. J. Sehni K. Greene W. M. Smith * E. W. Harrington R. L. Steller . R. P. Kenney R. W. Tannehill T. R. Linak R. L. Toole R. W. MacCuspie A. Vittone * J. F. Malone H. Waltemate * J. L. Nelson * B .M .G .Zwicker Steering Committee Members AKRON E . B. Katzenmeyer HENRY C.B.Cooper - M.D.Tawney LONG BEACH A.W.Clements - W.D.Robb LOUISVILLE P.H.Lawrence - P.A .Wagner S.S.Michels ITC (1) G. Huddleston F. Ells M. Mele ALTC R. L. Bowles L. Cohen E. A. Collins L. B. Crider (2) J. R. Goots R. M. Kreager B. K. Mikofalvy A. L. Schultz J. W. Summers J. A. TePas G. L. Wheelock HURICKTOWN A.R.Webber - J.W.Goetsch W. A. Reed ALTC Group Circulai D. E. Weaver G. J. Antifinger M. G. Morningstar G. D. Longeway R. S. Morgan D. E. Witenhafer oo C. A. Daniels J. A. Davidson L. Chandler J. B. Haehn H. H. Marty R. A. Jones H. H. Yieh W. S. Stebbins 24654009 BRECKSVILIE R. J. Fawcett R. A. Krueger D. E. Ley C.H.Lufter - J.B.Pausch ALGC L. V, Goode R.N.Rylands - R.S.Mather J. M. Whitney BRECKSVILIE Group Circulation (2) A. R. Berens T. R. Paxton R. H. Backderf T. W. Boyer R. J. Davis C. A. Marshall R. V. Kemp H. Kehe ALGC Group Circulation (l) R. C. Kaminski B. C. Knoble J. V. Laughlin BFG33329