Document 6bd28bBjOXEozNvyORQ14oMwo

CONFIDENTIAL EXTERNAL STEAM STRIPPING * OF PVC SLURRY January 18, 1991 Study by: Report by: G. W. Bacon R. J. Lahiere C. J. McDonald E. J. Meyer R. B. Newton P. C. Schirber B. L Tan A. J. Wiles Eric Meyer Pete Schirber VAB.0001167257 List of Tables, List of Figures...................................................................... Scope............................................................................................................... Executive Summary...................................................................................... CHAPTER 1: Regulatory Situation.......................................................... Current Regulations ............................................................................ Pending/Future Regulations ............................................................... PVC Industry Emission Levels............................................................ Conclusions on Regulatory Situation................................................. CHAPTER 2: External Steam Stripping (ESS) ..................................... What is ESS? ........................................................................................ Effect of ESS on VCM Emissions ..................................................... Effect of ESS on Resin Quality.......................................................... Impact on Operation............................................................................ Cost of ESS ........................................................................................... Debottlenecking Potential Resulting from ESS................................. Best Guesses, Assumptions, and the Like.......................................... CHAPTER 3: Alternatives to External Steam Stripping....................... External Stripping with Another Fluid.............................................. Stripping in a Fluid Bed Dryer .......................................................... Wet Cake Degassing............................................................................. Vent Collection and Treatment.......................................................... Improved In-Reactor Stripping .......................................................... CHAPTER 4: Economic Analysis............................................................ Base Case Economics.......................................................................... Sensitivities............................................................................................. Productivity and Cost Position vs. Competition .............................. CHAPTER 5: Licensor Evaluation .......................................................... Basic Technology ................................................................................. Specific Licensor Features................................................................... CHAPTER 6: Implementation Plan ........................................................ Reference List............................................................................................... Appendix ii 1 1 5 5 6 9 12 13 13 15 16 16 17 19 23 25 25 25 25 26 27 28 28 30 32 34 34 35 38 40 - - L VAB.0001167258 ,J-JV -**" i Ti'-h'Vw'lJiP't* < " -frjTt-qi I ' t . T nipii-- S-. - .. H `'i t T f 111- . .pit: LIST OF TABLES Table 1.1 Table 2.1 Table 2.2 Table 2.3 Table 2.4 Table 4.1 Table 4.2 Table 4.3 Table 4.4 Table 4.5 External Steam Stripping-Who has what technology VCM Stack Emissions External Steam Stripping Capital Costs Effect of ESS on 5385 Reactor Cycle Cost of Bottleneck Removal Debottlenecking Volumes and Margins Base Case Economics Economic Sensitivities Production Cost Comparison Reactor Productivity Comparison Table A.1 Table A.2 Table A.3 Table A.4 Effect of ESS on Aberdeen Reactor Capacity Effect of ESS on Aberdeen Plant Capacity Production Capacity - Oklahoma City Base Case Economic IRR Spreadsheet LIST OF FIGURES Figure 1.1 Figure 1.2 Figure 2.1 Figure 6.1 Figure 6.2 1988 VCM Emissions for all PVC Plants 1988 VCM Emissions for Suspension PVC Plants Simplified Flow Diagram For External Steam Stripping Forecast Capacity Utilization Timeline for External Steam Stripping/Debottlenecking Project Figure A.1 ESS/Debottnecking Economics: Projected Gross Margins and Payout versus Start-up Year A STUDY.ESS/PED4 11 1 --1H. |.1 .. . J..................-1. ,! 1- . .1 ^ ' v : i: -i t,:!.* * T >- -- ) t H'l'lfr . ; I -. ................ ii.... VAB.0001167259 SCOPE This report summarizes the team's work in evaluating external steam stripping of PVC slurry. Our scope of work was to: (1) define the need for external steam stripping, (2) determine its costs and benefits, (3) evaluate the alternatives to external steam stripping, and (4) identify and evaluate the licensors of stripping technology. A Based on these findings recommendations are made for "how" and "when" to implement controls to reduce VCM emissions. EXECUTIVE SUMMARY The primary reason for implementing external steam stripping is to reduce emissions of VCM. Currently, Vista's PVC plants rank among the highest emitters of VCM per pound of PVC produced. We are, however, well within the NESHAP regulatory level of 400 ppm. There are now two actions underway that may lower the VCM emission standard for our PVC plants. The first is the amended Clean Air Act that was signed into law in November 1990. The act calls for the EPA to issue standards on nearly 200 chemicals by the year 2000. The standards are to be based on Maximum Achievable Control Technology. No one knows for sure what level this will be for VCM. Current thinking in the industry is that PVC plants may not see a tougher VCM standard. If, however, the EPA were to adopt a lower standard, it is expected to be in the range of 25 to 50 ppm based on demonstrated performance by other PVC producers in some states. We expect that the EPA will act on the matter of a new VCM standard by the end of 1992. STUDY.ESS/PED4 1 ii 1 rrtl- ' -V VAB.0001167260 A The second regulatory action is at the state level. The State of Mississippi has adopted a policy requiring that risk to the local population be maintained below a specified level in order to obtain new operating permits. Based on some preliminary data, emission levels as low as 15 ppm may be required to meet this policy. However, there is still a great deal of work to do before we fully understand the state's policy and what actions will be required to comply. So far, the state of Oklahoma has not taken a similar position. The best method of achieving emission levels below 15 ppm is to steam strip the PVC slurry outside of the reactor in a continuous slurry stripping column. We call this process external steam stripping or ESS. ESS is capable of reducing emissions to a level of 1 to 5 ppm. This represents a reduction of over 95% versus levels of 100-250 ppm. All other resin properties are expected to be unaffected with the possible exception of "b" color which could improve slightly. The cost to implement ESS in both PVC resin plants is estimated at 12 to 24MM dollars. The technology is used by virtually every producer in the U.S., Europe and Japan. The technology may be purchased from one of four resin producers: Chisso, Norsk-Hydro, EVC, BFG. Purchasing the technology is the fastest, most cost effective method of obtaining the know-how. A benefit of ESS is that it provides the potential for significant debottlenecking. ESS replaces in-reactor stripping and thereby reduces the reactor cycle time considerably. To take advantage of this reduced cycle time, some bottlenecks in upstream and downstream equipment would need to be removed. We estimate that the capacity of Aberdeen could STUDY.ESS/PED4 2 1 A .r A\ Hf|l " "! Jt* !" " '-ftl : - -ii-1 i - - -1 VAB.0001167261 4 < -t j. r.. *................... . t m be increased by as much as 190MM lbs/yr at a capital cost of $10MM over and above the cost of ESS. Selling this additional capacity at the expected gross margin would provide a return in excess of 20% for the combined ESS/Debottlenecking project at Aberdeen. Further effort has been committed to define the opportunity for marketing this additional capacity and thereby take advantage of the debottlenecking potential associated with external steam stripping. Preliminary evaluation suggests that ESS/Debottlenecking is the best option for Aberdeen. A Another technology given serious consideration would collect the vents from the various emission sources. These vents would then be treated by incineration, carbon absorption or some other method to remove the VCM. Vent collection and treatment, however, is only capable of reducing emissions to the range of 20-50 ppm based on some limited testing. As such, it is probably not a long term option. The cost associated with vent collection and treatment is estimated at $2-5MM total for both plants. To the best of our knowledge, no one in the PVC industry uses this technology. Further work should be done to define just how much emissions could be lowered using this technology. A level of 10-20 ppm or lower could make this alternative attractive. This option has the most potential for OKC where debottlenecking is currently less attractive. Our recommendations are: (1) Proceed with design work on the portions of external steam stripping at Aberdeen that will require a long time to resolve. These portions are the dump system design STUDY.ESS/PED4 3 h r ..... . B-ih--. -f JWI ' 1- . :| -k? s U- ' 4 .-f -H' Ml T~ f |----- I -r. I VAB.0001167262 r ' -iTT p k-i-s-H'.r pr 11 .1: i.. ................. -1 s 1- j a: p. t and the recovery system design. The remaining portions of design work should be completed when regulatory requirements dictate implementation of ESS or when ESS is justified by additional resin capacity. (2) Evaluate whether debottlenecking Aberdeen as part of the ESS project is justified. If so, proceed with design work. (3) Complete work to define the potential for VCM emission reduction by vent collection and treatment. (4) Minimize VCM emissions at OKC through in-reactor steam stripping optimization. When future regulations require emission levels below the capability of in-reactor steam stripping, install either external steam strippers or a vent collection and treatment system based upon the regulatory requirements. STUD Y.ESS/PED4 VAB.0001167263 i-i i- y" l i - --ll i s -i - > ' <\ ri ' -si * vi* ' i i i| - i- . y.............. i. r... .................... . CHAPTER 1 REGULATORY SITUATION VCM emissions have been regulated for about 15 years. Over the past few years there has been renewed activity aimed at more stringent regulations covering VCM emissions. Vista must be prepared to make significant reductions in our current emission levels if tougher standards are enacted. The two sections that follow review today's regulations and what is coming in the near future. CURRENT REGULATIONS There are no problems meeting regulations currently in effect on VCM emissions to the atmosphere. These emissions are regulated predominantly by NESHAP which limits opening losses from the reactor. To meet this regulation, the plants must strip the VCM in the PVC slurry to a level less than 400 ppm. The plants have consistently met this requirement with typical results of 100 to 250 ppm. A OSHA also regulates VCM exposure to the worker at a level of 1 ppm over an 8 hour period. Current operating procedures allow us to meet this requirement routinely. At the state level, there are currently no regulations in force in Oklahoma or Mississippi with emission standards more stringent than the Federal level. However, some other states do have more stringent regulations on some PVC producers. B.F. Goodrich's plant in New Jersey, for example, is allowed the 400 ppm maximum on any given day but must average only 25 ppm over the course of the year. Their plant in Ohio is limited to a monthly STUDY.ESS/PED4 5 i H I'-ir I'l.,'*- I |-I^| . r.|i li: |i .... ij.i. .. l|I- ,. ,, p !.. j.,. !' i- " M -frf i -4 ' ^ i r r> . > ivi . ttf- VAB.0001167264 11 . ij u -.ivi i - -i- M -- A average of 200 ppm. Formosa's new plant is Baton Rouge has a permit for .5Q ppm on a quarterly average while Georgia Gulf in Louisiana has a permit limit of 50 ppm as a daily maximum. Similarly, PVC plants in Texas must meet lower limits (45 ppm on an annual basis for Formosa and 130 ppm for Oxy). The significance of these lower limits in some states is that a precedent has been set that Mississippi and Oklahoma can look to when considering renewal of permits. Furthermore, a standard of sorts has been set for potential reductions in the Federally regulated level. If the EPA considers revising the VCM emission level for PVC plants, it will be able to point to the BFG plant's permit level of 25 ppm as a benchmark. PENDING/FUTURE REGULATIONS Two significant actions are underway that could affect the VCM emission limits. The first is the Clean Air Act amendment which has just been signed into law. The second is a trend at the state level to evaluate acceptable emission levels based on perceived risk to the population. The Clean Air Act which was signed into law in November 1990 will require Maximum Achievable .Control Technology or MACT to control emissions of 189 air toxics including VCM. MACT will most likely not specify the exact technology to be used to control VCM emissions but rather dictate what level of emissions must be achieved. It is not clear what level this will be for the PVC industry. Current thinking is that the industry may not see any reduction in the emission standard for two reasons. First, the industry has been effectively STUDY.ESS/PED4 6 h hi- r : -b f - "* VAB.0001167265 r t. * regulated for the past 15 years and does not need further action. Second, batch processes including PVC production may be exempted from the new standards. If, however, they do decide that a new standard is warranted, it is expected to be in the range of 25 to 50 ppm. Timing for compliance is not yet defined for each of the 189 regulated chemicals. Standards must be promulgated for all chemicals within 10 years with 40 of these being completed within 2 years. VCM is expected to be in the first group. Compliance is then required 3 years after promulgation. Therefore, if VCM is regulated to a lower level, our probable deadline for compliance is the end of 1995. If MACT is not sufficient to reduce risk to health and environment, more stringent standards must be adopted within 8 years after the MACT standard is promulgated. For suspected carcinogens like VCM, these standards would require that the risk be no greater than 1 in 1,000,000 for the most exposed individual in the population. At the state level, Mississippi has recently asked the Aberdeen plant to meet VCM emission standards that may be more stringent than the Federal limit. The standard is based on a level of risk of cancer to the population surrounding the plant which is calculated using a standard formula and VCM levels at the plant perimeter. Fence line monitoring at the plant indicates that at some points VCM levels are 5 to 10 times higher than the level corresponding to a 1-in-10,000 risk. Further work is underway to better understand and quantify the actual risk to the local population as well as understand what VCM sources STUDY.ESS/PED4 7 rf: . hWf-k- i " " *- ' . r i j - - .* VAB.0001167266 L' , i-i --h.i -i-s-t . i i - ,* , > A affect fence line concentration the most. Until this work is completed, we will not know what role ESS might play in meeting this proposed state standard. The state has also said that ultimately we may need to reduce the risk to a level of 1-in-1,000,000. The plant's permit is scheduled for renewal in March of 1991. STUD Y. ESS/PED4 i - r i- ij -iii - . :i .l Mi..-ii1.i .................... i-i. - t i . j.-I. k k- T-f'1-4- (W'n i.'in -r-* --iru'K-jril i.-1 I _| ^ ^ VAB.0001167267 yr . . hi- -i -"i > H - . I . EMISSION LEVELS (Or who will With the advent of mandatory reporting of VCM emissions under SARA 3.13, it is now relatively easy to determine where Vista stands in the industry. Taking the PVC industry as a whole, Vista is placed in the upper half for emissions but still well below the highest two producers. In other words, we don't stand out from the pack. This is illustrated in Figure 1.1. Figure 1.1 1988 VCM Emissions For All PVC Plants S-$upenion 0-Disprion Occidental, PA S/O Georgia Gulf, OE S/D Borden, IL S/D Goodrich, NJ S/D Goodrich, OH S/D VISTA, OK S VISTA, MS S Goodrich, IL S/D Goodrich, LA S Borden, LA S Occidental, LA S Air Prod., KY S Shtntech, TX S Occidental, TX S Air Prod, FL S Goodrich, KY S Georgia Gulf,LA S Keyaor, CA S Goodrich, TX S 0 200 400 600 800 1000 1200 LBS VCM/MM LBS PVC A Emissions in Figure 1.1 represent the sum of fugitive and stack emission sources. Stack emissions are those emissions from definable point sources like blend tanks, dryers and silos. Stack emissions are directly related to the extent of VCM removal from the PVC slurry prior to dumping it to a system which is vented to the atmosphere. STUDY.ESS/PED4 VAB.0001167268 .r - V - "V -'l .r. , KFfl'lf'l ' 'J.-1 - * N1 1 ' ' 1 ti A While we do not stand out when viewing the PVC industry as a whole, a different picture appears when we look at those locations where PVC is made by only the suspension process. The suspension process is used to make over 90% of the PVC in the U.S. Among the suspension PVC producers, Vista's plants stand out as the number one and two sites based on stack and total emissions. This is shown in Figure 1.2. Figure 1.2 1988 VCM Emissions For Suspension PVC Plants VISTA, OK VISTA, MS BFG, LA Borden, LA Oxy, LA Air Prod., KY Shintech, TX Oxy, TX Air Prod., PL BFG., KY G. G., LA Keyeor, CA BFG.. TX 0 25 50 75 100 125 150 175 200 LBS VCM/MM LBS PVC STUDY.ESS/PED4 |. . .i ii . - r r I- " VAB.0001167269 'V-s j- ........................................... ,. .(> ''I'lit'flf................... A The reason we stand out from the competition is due to our method of removing VCM from the slurry. Virtually every other PVC producer removes VCM from the slurry in an external steam stripping column. External steam stripping is capable of reaching lower RVCM levels than in-reactor stripping because of better steam-liquid contact within the column. The table below summarizes what we know about other PVC producer's technology for removing VCM. TABLE 1.1 EXTERNAL STEAM STRIPPING-WHO HAS WHAT TECHNOLOGY PVC PLANT LOCATION TECHNOLOGY Aiscondel Atochem BASF BF Goodrich Borden Chisso EVC Formosa Georgia-Gulf Hoechst Keysor Norsk Hydro Occidental PW Resins Solvay Shintech Sumitomo Wacker Vista Spain France Germany Avon Lake, OH Deer Park, TX Henry, IL Louisville, KY Petricktown, NJ Plaquemine, LA Geismar, LA Illioplolis, IL Japan Europe Pt. Comfort, TX Baton Rouge, LA Plaquemine, LA Germany Saugus, CA Norway/UK Baton Rouge, LA Burlington, NJ Pasadena, TX Calvert City, KY Pensacola, FL Europe Freeport, TX Japan Germany Aberdeen, MS Oklahoma City, OK Under Development Atochem In-Rx, plus BASF wet cake stripper BFG BFG BFG BFG BFG BFG BFG 9 Chisso Converting to EVC from BFG Chisso, Formosa Chisso BFG Hoechst 9 * Norsk Nydro Shin Etsu 9 Shin Etsu BFG BFG Solvay Shin Etsu Sumitomo Chisso In-Rx In-Rx STUDY.ESS/PED4 11 VAB.0001167270 t if1 ^ -it- > A Unfortunately the regulatory picture is not very clear right now. The amended Clean Air Act, according to current thinking, will not likely result in tougher emission standards in the next 10 years. The State of Mississippi will require the Aberdeen plant to meet more stringent standards based on perceived risk to the local population. However, it is too soon to tell what modifications, if any, will be required to meet the new standard. We do know that other states are regulating some of our competitors to levels well below the NESHAP standard. Furthermore, we know that virtually all of the other suspension PVC producers have substantially lower VCM emissions due to their use of external steam stripping systems. Ultimately, we will need to reduce our VCM emission level. The time frame when we will need to make this reduction is not known and probably won't be for a year or more. STUDY.ESS/PED4 VAB.0001167271 i* .................. . ............... k M'l n . - i ^ [ r- -I- : - r i -1 - r- - i - erf -h-h. i > i . 'Apt |p.>.l I . r. r.. 'uui * ..... . . . CHAPTER 2 A Because of its ability to reduce residual VCM levels in the slurry to very low levels, external steam stripping (ESS) appears to be the best long term solution to meeting future regulatory requirements. This section describes what ESS is and what VCM emission level it can achieve, what effect it has on resin quality and plant operations, what it will cost, and what debottlenecking potential it provides. WHAT IS ESS? Currently, both the Oklahoma City plant and the Aberdeen plant remove residual VCM by in-reactor stripping. After polymerization is complete the reactor is depressured to a VCM recovery system. Once depressured, a vacuum is drawn and steam is injected into the reactor. The water vapor and residual VCM are drawn from the reactor and reclaimed in the VCM recovery system. This entire process takes 45 to 115 minutes of a total batch time of 320 to 460 minutes depending on reactor size and resin type. VCM levels in the slurry are reduced to a typical range of 100-250 ppm. With external steam stripping, the slurry is dumped at the end of polymerization through an enclosed dump system to a dump surge tank. (See Figure 2.1) The slurry is fed continuously from the dump tank through a feed-bottoms exchanger to the top of a stripping column. The slurry flows down through the column in counter-current contact with steam injected into the bottom. Specially designed sieve trays are used to provide effective contact while minimizing resin overheating. The stripped slurry from the bottom of the column is STUDY.ESS/PED4 13 VAB.0001167272 I- r I- F. -h 'Mft ' 1 -tn-f-f-t rn IT |l K-'-Ci i- i j; | f; |H| ' to- I -rriVr.-*!: torii'i 1-1." .S-.1 i I ta -1 --m A FIGURE 2.1 SIMPLIFIED FLOW DIAGRAM FOR EXTERNAL STEAM STRIPPING . i .i , . ix .1. , i . r .11 Ji i tf .f h .. : I-,. , . .. >p i . i- -- - - - -rfr.,--.. 14 VAB.OOOl167273 "HI -f: i"| P "|'C: p -.........................- +. .. i. -h i[.:: . , xi,. i ,i - i-|^. lm sent back through the feed-bottoms exchanger to one of the existing blend tanks. The vapor from the top is cooled to condense water for return to the column while the VCM-rich vapor is sent to the recovery system. A Because Aberdeen routinely runs three resin grades at a time, three complete external steam stripping systems are needed. At OKC where only one resin grade is made, one complete ESS system is needed. However because of design capacity limitations it is likely that two stripping columns will actually be required. Additional details on the design of the ESS systems are presented in a separate report (Ref. 1) EFFECT OF ESS ON VCM EMISSIONS Information from licensor claims, patent literature, and pilot scale lab testing (Ref. 3) all indicates that ESS will reduce slurry RVCM to 1 to 5 ppm without affecting resin quality. The lower porosity of low molecular weight resins will make it difficult to strip much below 5 ppm and still maintain good resin quality. However, 1 ppm or less should be achievable on the more porous high molecular weight resins. By comparison, we now average 100-250 ppm. The table below summarizes the effect of ESS. TABLE 2.1 VCM STACK EMISSIONS (LB VCM/MM LB PVC) CURRENT ESS % REDUCTION Low MW 260 5 98 STUD Y.ESS/PED4 ABERDEEN 5385 5415 150 115 11 99 99 TOTAL 163 2 98 OKC 5385 188 1 99 15 t i-'l I |t. - >, 4- h VAB.0001167274 > - |- k "h "k * A Converting all resin grades to ESS reduces emissions by over 98%. Operating at this reduced level would place Vista as the lowest emitter of VCM per pound of resin among U.S. producers based on 1988 figures. EFFECT OF ESS ON RESIN QUALITY Tests are being conducted with two potential licensors of ESS technology to determine the impact on resin quality. Results of work done so far indicate that there will be little, if any, effect from external steam stripping. Two potential licensors, Chisso and Norsk-Hydro, have conducted pilot steam stripping tests of our 5305 and 5415 resins. The resin samples were analyzed before and after stripping to determine the effect on color, heat stability, cold plasticizer absorption, and particle size. The results (Ref. 5) indicate that there will be little change in any of these properties with the possible exception of color. Although inconclusive, some of the test data suggest we might see some improvement in "b" color. We are also analyzing resin samples from Chisso and Norsk-Hydro's regular production. Results will be reported separately when all the tests have been completed. IMPACT ON OPERATION Installation of ESS at Aberdeen will limit the plant to producing only 3 resin grades at one time compared to the 4 grades they sometimes produce simultaneously today. The additional equipment associated with external steam stripping will also extend product switches as more time will be needed for cleanout. STUDY.ESS/PED4 16 VAB.0001167275 t l-Tl ri The product runs will need to be campaigned at a rate near the design capacity of the strippers. This means that the plant will not be able to routinely operate only one reactor on a resin grade (since the strippers will be designed to handle slurry from 3 to 4 reactors). It also means that if demand for a given product exceeds the capacity of one column, the plant will need to switch a second column to that product. They will then be limited to producing only two products at the same time. The plant will still be able to handle the occasional coarse particle size batch. To do this, the batch will be stripped in-reactor and sent directly to an off-spec blend tank. A It should be noted that these limitations in operating flexibility are typical of what the rest of the industry is living with. Also, these limitations would have minimal effect on operating costs. Oklahoma City will not experience any major changes to their operation if ESS were implemented there. COST OF ESS The capital costs for implementing ESS in Aberdeen and OKC were estimated by GED (Ref. 2,3,4) based on preliminary designs by PED (Ref. 1). Preliminary estimates are considered accurate to within 25% for the scope of work being estimated. However, because the scope is still not fully defined, the actual accuracy of the estimate is more likely 35%. The costs are summarized on the next page. STUDY.ESS/PED4 -+ mi' v. - i i. . VAB.0001167276 -jp-'T-k r. -I?" :: 11. 1 n- rf-k'ljl*. ; `tJPfp............ I" -*- - .................... - ii^i -- i- -j "I ** 1J' ' II t .i#':--*! ' :-iiri ; ii|'i . r x>.f - HI P-n-li . s. ; TABLE 2.2 EXTERNAL STEAM STRIPPING CAPITAL COSTS ABD Low MW System ABD 5385 System ABD 5415 System TOTAL ABERDEEN OKC 5385 TOTAL PLANTS LICENSING FEES (50%) GRAND TOTAL These ESTIMATED COST ($MM) 4.4 3.5 3.5 11.4 5.5 16.9 1.5 18.4 35% COST RANGE ($MM) 2.9-5.9 2.3-4.7 2.3-4.7 7.4-15.4 3.6-7.4 11.0-22.8 1.0-2.0 12.0-24.8 A STUDY.ESS/PED4 --Ir + i .k- i 'TP--.- +. i- .'F4fl: VAB.0001167277 DEBOTTLENECKING POTENTIAL RESULTING FROM ESS Implementation of ESS has the potential to increase Vista's total resin production capacity by as much as 270MM lbs/year or 33%. By removing the stripping process from the reactor and placing it in an external stripping column, batch cycle times are reduced by 30 to 90 minutes depending upon resin grade and reactor size. In addition, the need for evacuation is eliminated since the process is totally closed. This cuts the cycle time by another 6 to 10 minutes. Furthermore, installation of the clean wall technology has eliminated the need for chem wash and the second water rinse. Current plans call for this cycle time savings to be offset by additional reactor steam stripping to reduce emissions. However, with external steam stripping this time savings could be captured resulting in a reduction of approximately 12 to 20 minutes per batch. In total, 48 to 120 minutes can be eliminated from the batch cycle. A Finally, because stripping would take place outside of the reactor, larger batch sizes may be possible without having the problem of resin carry-over into the VCM recovery system. We estimate that batch size could be increased by 10% at Aberdeen and 5% at OKC. The combined effect of these cycle time reductions and increases in batch size results in the potential for an additional 180 to 190MM lbs/year of resin capacity at Aberdeen and an additional 80 to 85MM lbs/year of resin capacity at OKC. An example of the effect of an ESS project on reactor cycle is provided in Table 2.3 below using 5385 production in Aberdeen. A complete breakdown of all resin grades and the STUDY.ESS/PED4 19 VAB.0001167278 * 'll -t.......................r r-' pi. `i .1:31 . - I ; ` .v^t.'-i u 1 .r..= r- p. im: -E-f jj-.- l--,i -ii m. ? .. ,r , 'i* -`-..-Ti-'.r- JV 1i- :( r r-1nS-.lv------------- . ' - A effect of product mix on the capacity increase are detailed in Tables A-l, A-2 and A-3 found in the appendix of this report. TABLE 2.3 Effect of ESS on 5385 Reactor Cycle STEP Evacuation Charge Poly Recovery Dump/Rinse Chemwash/Rinse Cleanwall Slack TOTAL APRIL '90 (Minutes) 10 10 248 71 35 35 0 28 437 After Debottlenecking^1* (Minutes) 0 11 248 0 50 0 15 0 324 (1) includes effects of cleanwall, DCS, ESS, and larger batch size To take advantage of this increase in reactor capacity, bottlenecks upstream and downstream will need to be removed. Some preliminary work was done to identify the bottlenecks and the cost to remove them. (Ref. 6). The results, summarized in Table 2.4, show that a capital investment of approximately $10MM is required. The $10MM capital investment at Aberdeen to expand capacity by 190MM lb/yr represents a cost of 5.3 cents per pound of additional annual capacity. By comparison, adding a new module or building a new plant would cost 10 to 30 cents per pound of annual capacity. STUDY.ESS/PED4 20 J -t '.k 1 n -KV .|i. 'irtn ' V K il-,!' M'-r.'.:. VAB.0001167279 'll - il<4 r if-'* A TABLE 2.4 COST OF BOTTLENECK REMOVAL AREA ABERDEEN ( +190mm lbs/yr) Revision OKC (+84MM lbs/yr) Revision VCM Unloading 2 new spots VCM Storage OK VCM Transfer OK Blend Tanks OK Suspending Agent System OK Dryers new fluid bed dryer. increase rotary dryer rates Silos Extend height on 14 silos Utilities add one cooling tower Wastewater Treatment OK Staffing 1 operator/ shift for rail< car loading, etc. Other .'V. 4 4 14*1 P tI I4 , . *-.*%** I h I* 4 * 4 4 14 * m 4 # + * 4 i 4 ft 4 + lib 4 4 ft 4 d P - i. 4 i.h 4 4 14 4444 h 4.1 P * I P OK * F_ "'//('i'. "^* 4 'i' % 4% + i 4 444 P44 PI * P * 4 P 4 I P 4 KJ P 4 14 P I .4 P * 4 PI * I 14 4 1 44 14 14 4 4 + * 4 4 4 4 4 4 4 4 4 " 4 r 4 pJf p I f P + | * P 4 4 41 4 44 4 4 1 4 4 44 4 I 4 I 4 + 4 PIP 4PI4P4RIV 41 fl 444 444 4 4 + 4 * I 1-4 4.# -4 4 4 1 4 41 414 4 r ...................................' I I * 4 1 |_4 I 44 44 4 4 4 44 +4 P+ 14**41 m. P P 4 I- P.-P < P *.l ** P *. + . * b 4 1 P -4 4 I 44 P 44 P 4 .4 4 4 4 14444 P + P 4 ft 4PP44P44P44 P 4 * I P44P4 4PP + + P4d I V 4 P 4 PP44P4 + 4 + . _ PPP4PP4I P 4 4 4 *.^>4 4 4 - _ + + pppp + 4 Z4 v * 4 v 4 * _4_"_M_4. : w_ +* .*>"** P P f + * P_ >_ i_ > +44V 4 4 P 4 <4 4 P_ + _' . * +. P 44 P I V 4 P 44 Ip * A p + Vi'.-.-V J '.f $MM/yr OK OK OK OK OK OK $MM/yr add 2 silos add one cooling tower OK 2 operators/ shift see note 1 i < p + ..** * F 1 4 l * + h + ' 4 p1 p .p -p -p p -p. + _p p _p + 4 I 4f h P i * + 14 4 + 4 4 *_ a 44P+ + 4 4P P 4 4 PI P44P + I P d 4 + + * P+ P 4 4+4 + I -I H + 14 + + - I * t* I i d ft f 4 4 +14 * 4*4 414 b 14b h P + 4 444 44 44 h4 4 4 14 4 4 44 4 b 4444+* b+++ . " - -- "4 4 4 b 4 b + 4 144+44 I 4 + 4 144 14 ' + + + + + P 4 4 + P4 b i * p i ft fll-FII-ill-i + 44 P4 + 4P + I 44 1+ P # _P 4 * p .+ * 4.4.4 I ft I b ft ft<ft*l'ft,ft>#^ft,| b+#4d 44d bd a pA +44 + i p P44 + I R t 4 + p i i ft b p f l + P P + P4 + P 1 f ft ! 4 4 P 4 4 + 4 h + P + I 144 PPP4 + P4+P44P* J ft + PRP4PP4PPH I h ft ft * b * b P + P + 4 *p-p d. * ft +I PP 44 I d NOTE 1: Reduced cycle time will require more operators and/or improved automation of the process. This has not been defined. For the purposes of this study it was assumed that added capital to improve automation would be done. STUDY.ESS/PED4 i .1- T hi !* -rir VAB.0001167280 Debottlenecking upstream and downstream equipment does not necessarily need to be done at the same time as implementing ESS. From a design standpoint the ESS equipment could be sized for the expanded volume so that in the future the plant could be debottlenecked without replacing these systems. To operate the columns at today's rate, additional water would have to be added to the PVC slurry feed. This would result in an additional energy and water use of $100M/year. Another factor though is a regulatory one. The trend today is to require plants to reduce emissions when expanding capacity. If ESS has already been installed, there will be little opportunity to further reduce emissions to offset an expansion of 40%. However, there may be ways of working with the state to ensure that the plant is properly ''credited" for the emission reduction resulting from ESS so it can be applied toward a debottlenecking project in the not-too-distant future. STUDY.ESS/PED4 H- |. , -i-v.- .. f . t II - r^i+' L VAB.0001167281 I1-! F'i f H ^ iJ IJ r? < HU ---------.------------hh............ A While developing our findings on ESS, we ran into some questions which could not be fully addressed within the scope of this work. In these cases, a preliminary assessment was made using our best judgment and experience. These "gray areas" are discussed below: Larger Batch Size - The estimated increase in batch size of 10% at Aberdeen and 5% at OKC are based on some information we have obtained from European and Japanese PVC producers. The information shows some reactors operating at 90% full versus 75-80% at OKC and Aberdeen. We estimate that a 5-10% increase in batch size is possible since there will be less of a problem with foaming and resin carry-over due to VCM recovery in the reactor. However, batch size is not only influenced by VCM recovery problems but also by agitation. This is especially true in the extra large reactors in Aberdeen. Increasing batch sizes may cause a stagnant area at the top of the liquid which would result in poor quality resin being produced. Improvements in the agitation system could possibly alleviate this problem. If larger batch sizes are not achieved, the capacity increase at Aberdeen would be 145MM lb/yr (vs. 190MM Ib/yr) while at OKC the increase would b ih/vr (v* r 80-85MM lb/yr). Recovery System - The recovery systems in Aberdeen and OKC are belie_____________t___ to handle the additional VCM load after debottlenecking. This is possible because of two reasons. First, the instantaneous load to the recovery systems is not expected to change. STUDY.ESS/PED4 "23 / VAB.0001167282 A Second, the amount of time per batch when a reactor is tied to the recovery system decreases. In the current system the reactor must be depressured and then steam stripped through the recovery system. Only one reactor can be utilizing the recovery system at any time. This period lasts for 40 to 85 minutes for each batch at Aberdeen. With ESS, the reactor will only need to be depressured, and not steam stripped, through the recovery system. Therefore, each reactor will be using the recovery system only 30-45 minutes during each cycle. This reduced period of recovery system use will offset the shorter cycle times. While we believe that the recovery systems are adequately sized to handle the resin capacity increase, further work is needed toconfirm this. If additional recovery system capacity is needed, the cost for debottleneckingScould go up by Several million dollars./ 11 STUDY.ESS/PED4 r- -I ..............- .. --. . .. h VAB.0001167283 r ' t' - ' v u-ff-"-. 'M < !-" ............ i-- - ..........................._..............................,4p-.-nv* " rJ HPUjh-I,. Sfr J| * PI ..................................... .>|HtnHl.-rT . . CHAPTER 3 TO EXTERNAL STEAM STRIPPING Several alternatives to the proposed ESS project were evaluated. Of these, only Vent Collection and Treatment appears to have some potential. A brief discussion on each alternative follows. A more detailed discussion is documented in a separate report. (Ref. 7). A External Stripping With Another Fluid - This alternative is similar to steam stripping except another fluid, such as nitrogen, is used to extract the VCM. The advantage is that improved energy efficiency might be possible since the gas does not need to be vaporized and condensed. However, separation of VCM from the gas would be difficult. The cost is expected to be substantially higher than ESS. Stripping in a Fluid Bed Dryer - In this case, a new closed dump system would be used to contain the VCM in the slurry. The centrifuges would also be "closed". Nitrogen would be used in a fluid bed dryer to both reduce moisture and remove VCM. Water and VCM would be separated from the nitrogen via extraction, absorption or membranes. This alternative is more costly and unproven. Wet Cake Degasring - In this option, slurry would still be stripped in the reactors. The existing dump system and centrifuges would be closed to contain the residual VCM. The wetcake from the centrifuge would be passed through a steam degassing unit similar to those STUDY.ESS/PED4 25 ............................ ifj'i i . ^1. i ^r j r t i ii w-r i t r' ............................. k VAB.0001167284 used by BASF before going to the dryers. While being slightly less expensive than ESS, it has three disadvantages. It does not offer debottlenecking capability, it may not produce low enough RVCM levels in the resin, and it appears to be more likely to result in burnt resin. Vent Collection and Treatment - In this option, slurry is stripped in the reactor as is currently done. The existing dump system and blend tanks would be enclosed to collect the vents. The vents would then be treated to remove the VCM by carbon absorption, UV or catalytic oxidation, membrane technology or incineration. This alternative would result in RVCM levels in the range of 20 to 50 ppm. The total cost for both plants is estimated at $2-5MM. This appears to be an attractive alternative to ESS depending upon certain factors. The first factor is what level the MACT standard for VCM turns out to be. If the standard is 50 ppm, this technology is feasible. If 20-40 ppm, optimization of this technology will be needed to make it practical. Even if optimized, however, it is unclear whether this level will be sufficient to produce acceptable fence line concentrations at the Aberdeen plant. Another factor is that vent collection and treatment offers no potential for debottlenecking and therefore, no opportunity to improve fixed cost efficiencies or expand resin production capacity. For these reasons, this alternative is not recommended for Aberdeen. It may, however, be an acceptable method of reducing emissions at OKC where additional capacity is not currently attractive. The cost for OKC would be only $0.5-1.5MM. Further work needs to be done to determine just how low of a VCM emission level is possible using this STUDY.ESS/PED4 I- r hi--' . s if.. t- VAB.0001167285 r . method. A more detailed evaluation of vent collection and treatment technology has been made. Findings are summarized in a separate report (Ref. 8). Improved In-Reactor Stripping - This method involves putting cooling water on the condenser during stripping to allow greater steam flow while maintaining the proper stripping temperature. This is being done to some extent at the plants already. It appears that VCM emissions levels of 50 to 100 ppm are possible using this technique. If so, this would reduce our emissions considerably but still leave us near the top of the list of PVC producers. It is therefore not a long term solution to meeting regulations that are expected to limit emissions to 25 ppm. STUDY.ESS/PED4 + j> J 1 -M-H- -1 II ff ^ r I .k LI j | r..i | I .fc-- : -- .-4 , 4- | fa. VAB.0001167286 Vi |*-r . I h 11 | I lr * -r-- -4 .1 p* I " 1 . CHAPTER 4 The primary reason for doing an external steam stripping project is to reduce emissions. However, as mentioned earlier, ESS provides an opportunity to expand our production capacity at reasonable costs. To determine if the sales of the incremental resin production would justify the cost of expansion, an economic analysis was done for a debottlenecking project at Aberdeen. The results of the analysis show that the combined ESS/Debottlenecking project does indeed have a good return, around 26% IRR, if the incremental resin capacity can be sold at forecasted margins. A sensitivity analysis also shows that timing the project to coincide with the upswing in the market cycle optimizes the return. By delaying the project to the bottom of the cycle, the return drops to 17%. Additional details are presented in the following discussion. BASE Base case economics were calculated based on the following: Timing - Capital expenditures were assumed to occur in 1993 with start-up in 1994. This timing would take maximum advantage of the resin business cycle based on forecasted margins. Revenues - Revenues were derived from sales of an incremental 190MM lb/year of resin from Aberdeen. Domestic sales were projected to grow from lOOMMlb of incremental resin in year 1 to 190MMlb in year 3 and after. Export sales were expected to make up the difference between domestic sales and the 190MM lbs/yr. Margins were forecasted based upon historical figures. Volumes and margins are summarized in Table 4.1. A STUDY.ESS/PED4 1 > -- it1 -- 'rsnr"' r - r i * - n - i. :1 - it - ' J1 d'li-.-'-.-i. >j* r i-i r.' t. u ..Hit.-... - 1V"'f -f-rTtrf in-r i- rrttn - i-i'->*.*.* r - " >- H - ' 28 VAB.0001167287 - I ........................ - -- -c--. i- Table 4.1 Debottlenecking Volumes and Margins Total Volume (MMlbs) Domestic Export Total YR1 YR2 YR 3 YR4 YR 5 YR 6 YR 7 YR 8 YR 9 YR 10 100 150 190 190 190 190 190 190 190 190 90 40 0 0 0 0 0 0 0 0 190 190 190 190 190 190 190 190 190 190 Margins (ff/lb) Domestic Export 15 9.0 12.0 9.0 6.0 5.2 4.8 4.6 5.1 6.1 3.0 45 15 45 15 0.7 0.3 0.1 0.6 1.6 These figures are, of course, preliminary. Work is underway to further define the potential customers, volume, and margins for the incremental capacity. Costs - A capital investment of $22MM for the combined ESS/debottlenecking scope was used based upon an estimate by GED of PED's preliminary design. Incremental fixed costs of $1.1MM were included to cover additional railcar leasing. Staffing costs were included to cover the addition of one plant operator per shift, one sales representative, one technical service person in 1993 and another in 1995, and one S&T analyst. Licensing costs totaling $1.2MM were capitalized. Escalation of 4% was assumed. A Results - Key financial results for the base case are presented in the following table. TABLE 4.2 BASE CASE ECONOMICS FACTOR IRR NPV @ 13% Disc. Payback Period Profitability Index RESULT 26% S11.2MM 4.2 Yrs. 1.49 STUDY.ESS/PED4 29 VAB.0001167288 'fll- u -r r. - I . .'J-J I 11 --Kr Ln. r - 4 - r. - 'l.-i," | r A SENSITIVITIES Sensitivities were run to evaluate the impact of potential variations in the base case assumption. Table 4.3 shows the results of the sensitivity analysis. A copy of the IRR spreadsheet is included in the appendix (A-4). TABLE 4.3 ECONOMIC SENSITIVITIES Base Case i Capital + 35% Capital - 35% Constant Costs Slower Sales Faster Sales (3) 4 Delayed Cycle Debottlenecking only (5) IRR % 26 17 43 21 15 30 17 60 NPV 1L2 4.0 18.4 8.4 1.1 13.6 5.2 22.4 PAYBACK PERIOD 4.2 6.0 3.2 6.0 5.1 3.8 8.5 2.8 P.I. 1.49 1.13 2.16 1.37 1.05 1.58 1.23 2.96 NOTES: (1) Assumes no escalation on costs and a fixed margin on resin sales of 6.6c/lb. (2) Assumes incremental resin sales of: Incremental Volume (MM lbs) Domestic Export Total YR 1 YR 2 YRS 3-10 75 100 115 90 190 190 100 90 190 (3) Assumes incremental domestic resin sales of 190MM lbs/yr in years 1-10. (4) Assumes the project is implemented at the "bottom" of the market cycle. The resulting margins are: STUDY.ESS/PED4 30 VAB.0001167289 v: -h * i . `Ki", i. - * ` ' <i -- - n--i r* hm^i" r `iffc j t 17 YEAR 1 2| 3 4 5 6 8 Margin cpp 6.0 5.2 4.8 4.6 5.1 6.1 | 7.5 9 9 10 12 9 See Figure A.1 for further details of payout versus start-up year. (5) Assumes ESS is a "must-do" project. Debottlenecking costs of $10MM would then yield these economic figures on sales of an incremental 190MM lbs/yr. STUDY.ESS/PED4 VAB.0001167290 . -H l iif l PRODUCTIVITY AND Debottlenecking the Aberdeen plant will improve its cost position and productivity to a will . I with increasing production capacity by 190MM lbs/yr with increases in costs of only $1.7MM per year, the cost of resin production improves by 1 cent per pound at Aberdeen. This is outlined below. TABLE 4.4 0* PRODUCTION COST COMPARISON ABD Production (MMlb) ABD Fixed Costs ($MM) ABD Ex-Plant Fixed Costs(l)<2) ($MM) ABD Corporate Fixed Costs(1) ($MM) ABD TOTAL Fixed Costs ($MM) ABD TOTAL Fixed Costs (cpp) Pre Debottlenecking Bdst Debottlenecking (FY90) / (FY90 Basis) 458 / 670 11.2 11.4 3.1 4.6 4.1 4.1 18.4 20.1 4.0 3.0 i cA (1) Based on allocating costs between ABD and OKC (2) Does not include offsite costs By reducing Aberdeen's total fixed costs by 1 cpp, the plant's cost of PVC production would be comparable to what we believe Occidental's and Shintech's to be. Aberdeen's cost would also compare favorably to the OKC plant's cost. STUDY.ESS/PED4 32 VAB.0001167291 r I -*' i- i' : -`r .i*fi r1 The productivity of each reactor at the Aberdeen plant would also improve significantly as a result of debottlenecking. Currently, the plant produces approximately 2200 lbs of PVC per gallon of reactor volume each year. After debottlenecking, this figure increases to 3045. By comparison, Formosa's new Baton Rouge PVC plant has a reported capacity of 850MM lb/yr in 8 reactors of 34,000 gallons each. This gives them a reactor productivity figure of 3125 lb/gal/yr. Most other domestic, European, and Japanese producers have productivity figures in the range of 2000 to 3500 lb/gal/yr. This information is summarized in Table 4.5 below. TABLE 4.5 REA R PR WTY PARI Producer Location Vista Vista Vista Formosa Shintech Aberdeen (w/ DCS) Aberdeen (post Debottlenecking) OKC Baton Rouge Freeport PW Resin PW Resin Oxy EVC EVC Chisso Atochem Atochem Pensacola Calvert City Pasadena Barry, Wales Wilhelmshaven. Germany Japan Balan, France Brignaud, France Capacity /Reactors (MM lb/yrY/ f 480 67cT 10 10 350 850 500 (in new module) 175 275 1200 220 330 6 8 4 7 4 12 4 6 350 10 260 8 -6 Total Volume (gallons) 220,000 Rx Productivity (lb/gal/yr) 2200 220,000 3045 113,000 272,000 140,000 3100 3125 3570 70,000 96,000 420,000 111,000 158,000 2500 2865 2860 2000 2090 140,000 85,000 110,000 2500 3100 1800-2400 F! L- !i I: ............................. I t J .< . L .11- I. .1 J p r--i ^ j l - .. h VAB.0001167292 ............r- -- p -I0P(V` ^ ! --i "i!'- r | . -v --III- - >- - I |F p- r- *- - - -> |I- ir 1* : !- urn** ':''n .e i. i-.- . CHAPTERS LICENSOR EVALUATION An evaluation of eleven potential licensors of steam stripping technology has narrowed the list to four companies: Chisso, Norsk-Hydro, EVC, and BFG. The basic technology offered by each of these four companies is similar in many respects. In addition, each company has certain unique features that will affect the decision on which one is ultimately chosen. BASIC TECHNOLOGY The features shared by all four technologies are: All employ a multi-tray stripping column to achieve counter-current contact between the steam and the sluny. All use some type of sieve tray design. All include as options the use of feed-bottoms exchangers to recover heat, and overhead condensers to reduce vapor recovery load. All operate near atmospheric pressure and temperatures of 200-240F. All accept slurry feed with 25-30% solids and 1-2% VCM. All claim to be capable of stripping to below 10 ppm VCM without affecting resin quality. All claim that the stripping units require minimal operator attention. The differences in the technologies are mostly related to the column dimensions and tray designs. * STUDY.ESS/PED4 -I . ' * - ' - * - j I *M' VAB.0001167293 SPECIFIC LICENSOR FEATURES 1. Chisso - The Chisso technology is used more extensively than perhaps any technology today with a total of 25 units in operation (15 inside and 10 outside the company). Chisso also has the widest range of scale-up experience having designed units from 544M lb/hr of PVC. They have a batch pilot facility and have been very willing to perform tests and provide technical information. The Chisso stripping column features trays with specially designed baffles which provide long hold-up, high efficiency, and narrow residence time distribution. The Chisso column is shorter than others because only a few of these high efficiency trays are required. Another unique feature is the use of water spray nozzles under each tray to prevent resin build-up. The main selling feature of the Chisso technology other than its proven performance is that it is economical. The estimated capital cost for implementing Chisso technology is in the low to mid range compared to cost estimates from other licensors. The estimates of steam and total utility costs for the unit are the lowest of any of the licensors. STUDY.ESS/PED4 t i I'--- VAB.0001167294 .......... ' 1 ...........................: <' II K! <!+ I- 2. Norsk-Hvdro - The Norsk-Hydro technology was considered the most promising in 1980 when an evaluation was done for the Lake Charles Grass Roots PVC Plant Project. Their technology and Chisso's were considered to be the front runners by Air Products (now PW Resin) when they were evaluating an alternative to the BFG technology they currently use. Norsk-Hydro has a very impressive glass pilot column and have been very eager to offer their technical services. The main selling point of the design is cleanliness. They operate nearly liquid full on each tray to keep all surfaces wetted. This reduces the chance for resin hang-up and burning that results in contamination and downtime. One concern is that the residence time in the Norsk-Hydro column is longer than average and may result in more resin discoloration. The estimated capital, operating, and licensing costs provided by Norsk-Hydro are all about average compared to other licensors. 3. EVC - The EVC technology was developed by ICI in Australia. EVC is implementing it in their plants and is only recently beginning to license it. EVC claims that their column requires very infrequent cleaning and allows for very quick product changes. The main drawback is that the technology is less proven than STUDY.ESS/PED4 r VAB.0001167295 M : r I- ............................................................................................................................ q.L-- some of the others. EVC does not have a pilot unit that could be used to verify the performance of their technology with Vista resins. A 4. BFG - The BFG technology is one of the oldest and by far the most used in the U.S. (all BFG plants and four licensees). They have also licensed to five companies outside the U.S. They have not licensed to any company in the U.S. for eight years and until very recently have been reluctant to work with Vista. The original BFG technology followed fairly conventional design practice for cross flow sieve tray columns (ie. typical tray spacing, weir height, and hole area). No special features were included to promote a narrow residence time distribution or prevent resin build-up. BFG claims to have made some design improvements recently. Additional information on the licensors is available in two separate reports (9,10). STUDY.ESS/PED4 r t |`-M - -t t -y <' I : 1 - '!J r: ............................ VAB.0001167296 ^ hi i lr i[ ' 'T' 1 "T ^^ CHAPTER 6 The ESS/Debottlenecking project will take approximately 3 years to implement. This includes 1-1 h years to execute the technology license and complete process designs and another lh-2 years to do the mechanical design and construction. The relatively long lead time required to implement this project means that we must closely monitor the timing of pending environmental regulations. We must also anticipate when additional resin capacity could be effectively utilized to take advantage of the debottlenecking potential. Currently, the business area forecasts show capacity utilization being relatively weak through 1993 but rising through 1994 and 1995. (See Figure 6.1) In order to have the additional capacity available to meet the rising demand in 1994, we would need to begin work on ESS/debottlenecking by January 1991. (See Figure 6.2) Figure 6.1 Forecast Capacity Utilization in the PVC Industry 1991-1999 STUDY.ESS/PED4 T .-I--- -I : -t If!. : 38 VAB.0001167297 *' ' ' l I " Hr Manpower requirements to implement an ESS/Debottlenecking project at Aberdeen are estimated at 60-72 manmonths of process engineering time. Project engineering would be largely handled under contract with Vista supervision. Some R&D assistance, perhaps 2-6 manmonths, may be needed from time to time throughout the project. Figure 6.2 Timeline for External Steam Stripping/ Debottlenecking Project A CY CY CY CY 91 92 93 94 Process Designs Licensing ' Estimating & AFE Mech. Designs Procurement Installation Commissioning & Start-up --- -- -------- STUDY.ESS/PED4 W-r- p-t 1 ft-- -tn- VAB.0001167298 T rs REFERENCES 1. Preliminary Design - External Steam Stripping. P. C. Schirber and B. I. Tan, 1/22/90. 2. GED Preliminary Cost Estimate - ESS Phase L Aberdeen Low MW. A. K. Weekley, 11/30/89. 3. GED Preliminary Cost Estimate - ESS Phase ILA Aberdeen Remainder. A. K. Weekley, 11/30/89. n/30/89. TM C ^ t ate ase IA. O a o a C ty, y, 5. IOC - Resin Quality Study - For Review. B. I. Tan, 10/1/90. 6. IOC - ESS Study. Plant Debottlenecking Requirements (Revised^ P. C. Schirber and B. I. Tan, 5/2/90. 7. IOC - ESS Study. Evaluation of Alternatives. P. C. Schirber, 8/11/89. 8. IOC - Vent Treatment System Technology Evaluation. B. I. Tan, 11/5/90. 10. IOC - ESS Study. Comparison of Stripping Column Designs (Revised!. P. C. Schirber, 9/12/89. STUDY.ESS/PED4 ... f ? *-f|: I -r ^1 3* ^(111 J". 4- i|-:.:C"! r .-.,..^1 I- | i!r .. VAB.0001167299 tlyr T rv k. II 1> T ! I I L. -k y - !n APPENDIX STUDY.ESS/PED4 r- h i-i' ip rv i VAB.OOOl167300 TABLE A-l EFFECT OF ESS ON ABERDEEN REACTOR CAPACITY REV . - 4/04/90 EVACUATION CHARGE POLY RECOVERY/STEAM STRIP DUMP/RINSE PURGE CHEMWASH/RINSE CLEANWALL SLACK TOTAL MINS STREAM FACTOR PVC/BATCH (MLB) AVGR. PROD. (MLB/HR/RXN) YEARLY PROD. (MMLB/YR/RXN) % INCREASE CUM X INCR. EVACUATION CHARGE POLY RECOVERY DUMP/RINSE PURGE CHEMWASH/RINSE CLEANWALL SLACK TOTAL MINS STREAM FACTOR PVC/BATCH (MLB) AVGR. PROO. (MLB/HR/RXN) YEARLY PROO. (MMLB/YR/RXN) % INCREASE CUM % INCR. EVACUATION CHARGE POLY RECOVERY DUMP/RINSE PURGE CHEMWASH/RINSE CLEANWALL SLACK TOTAL MINS STREAM FACTOR PVC/BATCH (MLB) AVGR. PROO. (MLB/HR/RXN) YEARLY PROD. (MMLB/YR/RXN) % INCREASE CUM % INCR. CURRENT 10 15 184 115 40 35 27 426 0.87 54.70 7.70 58.72 CURRENT 10 10 248 71 35 35 28 437 0.91 39.20 5.38 42.92 CURRENT 10 15 297 60 28 28 30 468 0.91 40.20 5.15 41.08 5305 IN 0-745 PLUS PLUS CLEANWALL DCS <1) 10 10 15 15 184 184 115 115 40 40 PLUS ESS PLUS LARGER BATCH < 15 184 60 (2) 17 184 66 15 15 27 0 15 15 00 406 379 274 281 0.87 54.70 8.08 61.61 4.9% 4.9% 0.87 54.70 8.66 66.00 7.1% 12.4% 0.87 54.70 11.98 91.29 38.3% 55.5% 0.87 60.17 12.83 97.74 7.1% 66.5% 5385 IN THE OLD MODULE PLUS PLUS CLEANWALL DCS (1) PLUS ESS PLUS LARGER BATCH (3) 10 10 10 10 248 248 71 71 35 35 10 248 45 (2) 11 248 50 15 15 28 0 15 15 00 417 389 318 324 0.91 39.20 5.64 44.98 4.8% 4.8% 0.91 39.20 6.04 48.19 7.1% 12.3% 0.91 39.20 7.39 58.94 22.3% 37.3% 0.91 43.12 7.99 63.73 8.1% 48.5% 5415 IN THE NEW MODULE PLUS PLUS CLEANWALL DCS (1) PLUS ESS PLUS LARGER BATCH (3) 10 10 15 15 297 297 60 60 28 28 15 297 45 (2) 17 297 50 15 30 445 0.91 40.20 5.42 43.21 5.2% 5.2% 15 0 415 0.91 40.20 5.81 46.29 7.1% 12.7% 15 0 372 0.91 40.20 6.48 51.64 11.5% 25.7% 15 0 378 0.91 44.22 7.01 55.90 8.3% 36.1% NOTES: (1) REDUCTION IN BATCH TIME THAT WILL RESULT FROM THE DCS PROJECT AS ESTIMATED BY QUAO "S'*. (2) EST. TIME FOR DUMP/RINSE WITH ESS DET. BY THE NEED TO AVOID OVERPRESS. THE DUMP TANKS. (3) POTENTIAL TO INCREASE BATCH SIZE BY AN ESTIMATED 10% W/0 IN-REACTOR STRIPPING. f" v - ................................... -.p.K ... -frf. rv .......................................................... ............................ .................. VAB.0001167301 -Xc XI 'I" *h 1 s 3 a * 0. "v * S i 0 OX m in <m NN K>0- Cm-K* fO- rvO3j m oo in *0* o*4 o4 m oo CM Kl CM m oo >o O*h m Kl m CM CM Kl o -r ImM ^O Ncm & $ tn >o o A a Kl Kl oo Kl CM CM OK n cm (M -- 3S 3 * A a. Ki m m Kl n-- Kl o m cm * o sm 4 Kl o o Kl M ^S 0 C--O Kl O cm cm I N> Ss CO K* Ml *- *- CM 1 3 > T *O CM fc K in sP O f CM CM CM 1 >o rvi 00 cO m CM ^ CM cO * >0 CM O *-- Kl ^ o o CM Kl * 'O K ^m> K) O 0* CM CM CM s 00 O (M N CM CM Kl in co - ki *- o* r* N4 0 Kl CM CM in o ^ ki ki *** f ^ IM ki *- ki Kml oomKl 4 t- o m m v* u o <N I < W </> J P3 U</J> < H 10 a. o CM Kl in Kl O 00 cm in i-- * 0 Kl sp Kl i O -- - cm m SiNnt m0 O >0 *CO O Kl *-- CM CM 00 Kl Kl - CM K* 4* W* r- CM CM CM Kl ^ Kl - O * 00 O^ -- *- CM 'O Kl Kl PO - CM * Kl 00 Kl CM - CM Kl CM O* CM *-- CM nO o m Kl O >0 ** Kl O O CM "4* Kl O Kl cm oo K> Kl m >0 CM Kl *- - CM 3 CM * - ms Mr*) C(MM CM- flO o o 3 :N* CM in > 0 1 <M cm o < m -- CM Sa ^ Kl cm no m 3 3 m cm o> m cm ^ s * 00 CM CP N NCM NO * O O 'Oo m 'O m m cm p O 'O o N- K. N** oom O O 00 4*4 CM - *- CM CM - CM CM *" CM CM Kl 9 gU 1/1 O CM 1 01 UJ 1 OC mm 3 oo Kl in m UcnJ CJ r CM Kl in 2 in in 3 00 Kl in in CM U0J CJ Kl Kl Kl CM NOTE: ( 1 ) TWO COLUMNS ON 5415 3 R-- *0 m 5 --J Ml m Kl U</J> CJ 2 _ m in 3R --i in m U</Jl CJ 2w mm 3S3 -j m m m UJ 0 CJ VAB.0001167302 i-..........................................................- f - ii-*|- = * -i f Hff-T' -Ut <S i- i -i' n- . - i 'i i. . it ; h >M- ''-Ifs-' .fH*.-- f .- : Pi...................... -HS^ 5. + I.:i Jir-.n," TABLE A-3 PRODUCTION CAPACITY OKLAHOMA CITY Evacuation Charge Polymerization Recovery Dump/Rinse Purge Chemwash/Rinse Clean Wall CURRENT WITH ESS 66 88 210 210 42 27 37 10 27 27 WITH LARGER BATCH 6 8.4 310 38.9 10 27 WITH CLEAN WALL 0 8.4 210 38.9 Stream Factor PVC/Batch (M lb) Avg. Prod. (Mlb/Hr/Rxn) Yearly Prod (MMlb/Yr/Rxn) Incremental % Increase Overall % Increase Plant Capacity 0.9 40.8 7.7 60.3 360 0.9 0.9 40.8 42.8 8.2 8.6 64.8 67.5 7.4 4.2 7.4 11.9 385 405 0.9 42.8 9.4 74.4 10.3 23.4 445 A STUDY.ESS/PED4 \ 'I - - 1' p - . ft - T t r L pi 'w '. ................................pit . VAB.0001167303 -< p :............. I.. :i t h | . rr I . 0I> c sI 4I <0 z 8 o g UJ CM D *- $ o99oo CM Si IgaggS oo o 3 51* oo 518r28 9 OO CM s d <o=! 3o 5 *CM CiSM N o Q 85 O 6 o99o " in o CM i-- 69 0 " *8?i? i|s i? s i e 1 | g g S3i8 S SsSSSai s ssagsgs CM CM 288 a s CM rv co *- CO oo oo oo oo oo CM 8 8 Sm88 8 O 25 -- CM -- 8 CM CM 8 O CM 8 8 oo Is i 83 * whJ P<HQ y 2 uo. *S8Iz 25 CO 5 88128 V- OV S8o Smo 2S3 8 5 fC 8 8 8 699'" oo co 8 a8 S S SS|^ (M t- r*. 3 8 O 8 CO 1 $ SI t- "n o o 8 3CO II I 9 8 O *1 8 8 88 J e <0 9 5 88 33 N^ CM *- 1- 1- 8 m 2 e z oa 0> lf C8M 8CM 8CM 8| $_ $_ II II CM CM I oo 8 UJ CO 5 UJ CO I UJ UJ fl8 UJ CD < *- 2 IIIIfOz *E a 5 a 5 Q UJ co tsifi 8 jfi<0 c .c > > 3 s 6 Iff hss 11 IIS tr r: 8 2 -9 </>*>, ^ IS o lii C Ql S>00=5 3 i II IS<o HillI 8 i<g <3 - .sa a -o c* c 6 is ia. uj C a ato 8 c o i O o 9 3 <0 *9C1Oc_) 8 *8 5 jc 3 IL. *- <3 9 at O C 3 C c 1 | *6E ^iI i8C < I g-3 a >o ^ O 3o S o c 8v* <^N o I 8 u * * sa 3 o 16 Ss3 A . . i: I- M- : . li -H-fr -t : ^h- -<*i- .if VAB.0001167304 s- k PVCGROSS MARGIN (VLB) 0.12 0.1 0.08 0.06 0.04 c I T' 1 u_ V 0.02 0 1980 1985 HIST. GM ^w .J A * ,, ESS/DBTLN ECONOMICS PROJECTED GROSS MARGINS 1990 OPTIMUM PRO. ECT LIFE 1995 BA PROJ. GM 2000 2005 2010 REP. CYCLE GM % IR R PAYOUT PERIOD (YRS) ESS/DBTLN ECONOMICS PAYOUT VS. START-UP YEAR 30 30 25 25 20 20 15 15 10 10 55 0 1994 1995 1996 1997 1998 1999 2000 START - UP YEAR 2001 %IRR PAYOUT PERIOD 2002 0 2003 VAB.0001167305 r p " p i - 1 ni j .i r