Document Oz9g5vv4MYJmVg6JGwROG2RR1

X c S] LOSS PREVENTION ... Monomer Storage and Protection Styrene and vinyt chloride are two major petrochemicals; both require care In handling to pre vent explosions and loss of quality. P. CL Shelley, end E. i. Sills, The Dow Chemical Co., Texas Div., Freeport, Tex. This article will discuss two monomers which are manufactured, stored and handled in large volume by the Dour Chemical Co. and others. These are styrene tad vinyl chloride. Styrene production in the United States in 1968 was about 8.6 billion lbs, and is expected to rise to 4.8 billion Sm. by 1972. Individual locations producing from 500 million to one billion lbs. of styrene monomer annually are not uncommon in the industry. Distribution from some of these plants is worldwide. Almost without ex* ceptfon, locations of major use are far removed from the producing points. This adds up to a lot of storage and handling. For example, Dow stores styrene at seven major locations in the United States and in per* bans ten foreign coentries. vfriyl chloride production In the United States in 1968 was about 2.8 billion lbs. and worldwide In excess of 10.0 billion lbs. Production is expected to rise in the United States to 4.5 billion Ibs/yr. and worldwide to 1S.0 billion by 1970. As with styrene, the location of VCM users is in most cases remote from the producing site, which makes large bulk storage and handling systems a nee* essaxy adjanet of the VCM production business. In many instances, some storage facilities at the user end are required as well. Unlike styrene, vinyl chloride is a gas at N.T.P., and therefore requires pressure storage under its own vapor pressure, or as an alternate, refrigerated vapor recorepression for low pressure storage systems. Despite these storage requirements, the growing trend is to distribute vinyl chloride on a worldwide basis. Styrene characteristics Styrene is s colorless non-corrosive, aromatic hydro* carbon liquid. Table 1 gives its more important prop erties aa related to storage and handling. In develI oping techniques and equipment for styrene storage I and handling, three major objectives, and a^ew minor I ones which will be developed, must be considered^ These | are personnel protection, protection of the quality and momerie status of the material under normal ci> | earnstaoecs and protection against fire. Personnel protection is accomplished primarily by I training and utilizing common sonse. Styrene is low In (single dose oral toxicity, but accidental ingestion cases {should be referred to a physician immediately. Vomitling should not be induced because of possible nsplraItion of the monomer into the lungs. Styrene in the eyes can be very painful, but the like lihood of permanent damage is fairly remote. Eye pro tection should be worn where the possibility of eye contact exists, as in sampling, transferring and main tenance activities. Washing contaminated eyes im mediately with water for 16 minutes is the recom mended treatment. Again, medical attention should be obtained as soon as possible. Casual skin contact normally causes little if any irritation. On the other hand, prolonged, contact, as would result from continually wearing clothing soaked with styrene, or frequently repeated contact is capable of causing blistering and possible skin swelling. Wash ing with soap and water is the remedy. No case of styrene absorption through the akin with resultant' systemic effects is on record. Table 1. Selected properties of styrene monomer. Chemical Formula Density, Ibs./gal. @ 77*F. Vapor Pressure, mm Hg @ 77*F. @ 104*F. Boiling Point *F. @ 760 mm Flash Point, Tag Closed Cup, *F. FTre Point, Tag Open Cup, *F. Explosive Limits in Air at Room Temperature % Viscosity, Centlpolse @ ?7*F. C,H,CH:CH, 7.8 6.5 16.0 293 94 99 1.1 to 6.1 .71 Table 2. Specification for styrene monomer. Property Sales Spec. Typical Anal, Purity, (by freezing point) % Minimum......................... Color APHA, Maximum ............ Saybolt, Minimum.......... . Aldehydes (as benzaldehyde), % Maximum by wt........ Peroxide (as HjOt) % Maximum by wt......... Sulfur (as S), % Maximum by wt.............. Chlorides (as Cl) % Maximum by wt........ Polymer Content Maximum--ppm........... 99.5 10 27 0.020 0.010 0.002S 0.01 10 99.73 <5 .004 .0007 .0001 .0001 0 63 AP06622001 Vapor concentrations below about 100 ppm in air are neither likely to be harmful nor particularly objec- tionable to most personnel. Concentrations above this level become increasingly more irritating to the nose and eyes. In cades of excessive exposures, an individual could be overcome, as by any other volatile hydrocarbon. Fresh air and rest should bring him back with little likelihood of permanent effects. As protection against possible leaks and spills, any styrene handling installation should be designed to minimize human exposure and provide adequate venti lation and drainage in critical operating sites. Person nel should be well instructed as to the nature and haz ards of the material and provided with safety goggles and ready access to a safety shower and eye bath. Styrene is marketed under the specification shown Table 3. Shelf life of styrene monomer. Effect of Inhibitor and Oxygen at Various Temperatures 12tolS ppmTBC SO to 75 ppm T8C Saturated with Less Titan Saturated with Temp. Oxygen 3 ppm Oxygen . Oxygen 60*F. 5 to 6 mo. 10 to 15 days More than 1 year SS*F. 1 tog mo. 4 to 5 days- 3to4months 110*F. 8 to 12 days Less than 24 hn. Lessthan30days Table 4. Selected properties of vinyl chloride monomer* Chemical Formula Boiling Point (760 mms.), *F. Melting Point, *F. Vapor Pressure at 77*F. Flash Point (Cleveland Open Cup), *F. Explosive Limits Vel. % in Air Liquid Density grams/ml. 77*F. --4*F. Liquid Viscosity cps 77*F. CHj.CHCl 7.2 -244.4 -- 56 pels --103.4 3.6-26.4 0.9013 0.9834 0.185 in Table 2. For comparaison, a typical analysis as man ufactured is included. Storage problems Polymer content and color are the most difficult qual ities to maintain in storage, especially polymer con tent Styrene has no significant value except as a poly mer or copolymer with other monomers. It Is in the latter field of use, especially when used as a co-monomer in the preparation of polyester resins, that small quantities of the horaopolymer adversely affect finished product quality. In addition to end use considerations, it is possible for the polymerization reaction, which is exothermic, to become self accelerating. This can re sult in an extremely dangerous situation, involving high temperature and pressure development in a con tainer, as well as loss of the product. In short, styrene monomer is a heat sensitive perishable commodity and must be recognized and treated as such. Polymerization in storage and shipping containers la controlled by adding an inhibitor. Although various compounds exhibit some degree of effectiveness, paratertiarybutyl catechol (TBC) is the one almost uni versally used. Minimum effective concentration for prolonged storage is approximately 10 ppm. TBC also acts as an antitloxidant and, strangely enough, is effec tive as a polymerization Inhibitor only In the presence of oxygen. Table 3 gives a rough indication of the TBC-oxygen system's effectiveness and clearly shows the effect of temperature on the system. Shelf life means the time the monomer can be ex pected to remain within specification. The table Is based on a combination of laboratory tests backed by many years of practical experience. Several conclu sions pertinent to quality protection in storage can be drawn from it: 1. The lower the temperature the better. 2. An ade quate concentration of TBC must be maintained. 3. An adequate concentration of oxygen must be maintained. Although not absolutely necessary, Dow considers It good practice to insulate and refrigerate styrene stor age in areas where average ambient temperature ex ceeds 80*F for a substantial period. Refrigeration is external and the monomer is circulated as required to maintain the body of liquid in the tank around 7675* F. Such an installation has the added advantage of tending to equalize day to night temperatures in the vapor space and thereby minimize monomer condensa tion on the tank's roof and sidewalls above the liquid level. TBC is graduaDy depleted from styrene in the course of doing its job as an inhibitor. Figure 1 shows ' this effect quantitatively. Concentration should never to allowed to fall below 10 ppm. System effectiveness can be maintained simply by adding TBC as required. Here again, means of cir culating the tank contents becomes important. TBC is added as s concentrated solution, about 15-20%, In monomer to achieve any desired increase in overall concentration. TBC content should be checked every few days in dead storage. Oxygen content is adequately maintained when sty rene is stored under air. The equilibrium concentration of oxygen in the liquid is around 50 ppm at room tem perature. Under air is always the simplest and gen erally most satisfactory way to store styrene where the tank is designed to minimize polymer accumulation on the roof; turnover is frequent enough to eliminate any problems arising from buildup of aldehydes and 84 AP00022002 I' \ ^t ,, :nr*w * / (s ^ *.r ***> Figure 2. Icicle growth in an air at mosphere on the unlined roof and supporting structure of a 70 ft. by 24 ft. high tank after 125 days service. S&Lft Figure 3. Horizontal steel tank measuring 10 ft. by 36 ft. in process of being cleaned after four years in styrene service. Figure 4. Top of epoxy coated 10 ft. by 36 ft. tank after six months in styrene service. Inspection after 3yj years showed no change. other oxygenated compounds. Shelf life as shows in Table 3 applies. Air storage problems Storing under air can. however, create problems as well as solve them. Since TBC is a high boiling com pound, monomer vapors above the liquid level in the tank are uninhibited. Normal temperature swings re sult In a continuous cycle of vaporisation and conden sation on any structure in the vapor space, including the roof and sidewalls of the tank. Droplets will adhere to any rough or porous surface, in the presence of air, these droplets polymerize readily into a highly dis colored and oxidized product. Polymer `'Icicles" develop and continuing refluxing action dissolves the material and carries it back into the bulk of theliquid, ifallowed to proceed indefinitely, these icicles will grow to tre mendous size'and'eventually fall into the tank, gen erally throwing the entire contents off specification. Figures 2 and 3 illustrate the condition. The situation described occurs to a harmful de cree only is uncoated steel tanks having internal roof supporting structures as shown. Rusty steel and com plicated structure seems to be the preferred environ ment for polymer buildup. A non-poreus smooth sur face provides no sites for retaining the uninhibited monomer and, even in the presence of air, little poly mer buildup can be expected. All modern styrene stor age to designed to provide such a surface within the bounds of reason. Self supporting dome roofs are em ployed. API 650 tanks having a dome radius equal to 0.8 the dia. of the tank are satisfactory, and can be constructed in sizes up to atleast 70 ft. in dia. The roof and eidewsQ down to within about 2 ft. of the bottom are coated with one of several suitable coatings. For large tanks, a coating which will cure at atmospheric temperature is the most practical. Several epoxy types fall within this category. The bottom and lower 2 ft. of wall is coated with a rust resisting inorganic zinc sili cate material such as the Dimetcote protective coatinga. This permits static charges in the liquid to drain off through the tank ground. Figure 4 demonstrates the coating effectiveness. In situations where it is necessary to continue using uncoated tanks for styrene storage, particularly those containing interna! structures, the best, solution to polymer formation in the vapor space istffpJd'the tank with an inert gas such as nitrogen or natural gas. The latter is usually cheaper and just as satisfactory. In the absence of oxygen, polymer formation is retarded to the extent that a tank may remain in service from five to ten years without cleaning. However, under this con dition, the oxygen in the liquid is rapidly depleted and polymerization will proceed, as indicated in Table 3. A serious quality problem is created unless the tank is subject to rapid turnover as, for example, a day tank. The solution is straightforward, but requires careful control if a flammable pad gas is used. Enough air is pumped into the liquid periodically or continuously to maintain a minimum of 10 ppm oxygen in the liquid phase. If styrene is to be stored within or near a chemical complex wherein halogens, particularly bromine, are used or produced, a tank breathing to the atmosphere creates an untenable situation with respect to air pol lution. Styrene concentrations in air as low as a few parts/billion will form an extremely potent lachrymator when contacted by bromine in sunlight. Only slightly higher concentrations will do the same thing with chlorine. Where this situation exists as, for ex ample, in Dow's Texas Div., all styrene tanks within rangeof a probable halogen containing atmosphere are vented through an independently fueled flare. Such precaution is necessary regardless of the storage at mosphere. Color development Color development in storage can be one of the most exasperating* quality problems with styrene. The reason is usually not immediately apparent and is sometimes never determined. One cause, discussed previously, is the extraction of color bodies from oxi dized polymer attached to structures within the tank. Another is a compound which sometimes is formed by toe reaction of TBC, moisture and iron oxide (rust). Still another is prolonged contact with copper-bearing alloys such as brass and bronze. Copper reacts with styrene to impart a characteristic blue-green color to the monomer, and should not be used In styrene han dling equipment. (All other common metals such as aluminum, stainless steel, galvanized steel, etc, are suitable.) Assuming the system is adequately designed to main tain overall quality as previously discussed, the best de fense against color problems is to avoid contamination by strict adherence to high cleanliness standards. For example, styrene which has been laying in a line for a w,eek or so, especially an exposed line, should never be flushed into a storage tank without prior inspection. If a batch of styrene turns up off-color it esn usually be restored by contact with activated alumina in a fixed-bed filter. TBC will also be removed and must be replaced immediately. Particulate matter can be extremely detrimental in some styrene end-uses, and is another reason for clean liness. it is standard practice to filter the monomer between bulk storage and shipping containers. Many users filter again immediately prior to use. Cartridge type filters, capable of removing 10 micron particles are commonly used. Figure 5 illustrates loading pumps and filters in a Dow plant. Note especially the canopy for shading. w TO 85 AP00022003 the liquid surface and also near tne center of the tank provides a particularly defirnble situation. The test tank was approximately 11 ft. in dia. and contained 6 ft 3 in. (about 3.400 gal.) of styrene. The i foam used wits National Foam Systems, Inc. Aero-O1 Foam XL3 liquid, a foam containing tluorinated com pounds. The foam whs applied in 4% concentration and admitted to the center of the tank bottom, having passed through an aeration device or "foam maker.'1 Foam was started 5 min. and 15 sec. after light-off at 1t Figure 5. Styrene loading pumps and filters. a rate of approximately 0.1 gal./min./sq. ft of liquid surface. Results of this test are tabulated below: TIME DESCRIPTION 0*0" ............. Light off SIB*............. Foam Injection started(valve opened) l t Although styrene is classified ms m reactive flam* mable liquid, it is a relatively low hazard material with 5*32"............ Sound of foam on hot tank walls 6'0" ............. Fire very low-mostly around edges respect to fire protection. Figure 9 is a plot showing 6'20"............. Fire around edge only the range of temperature and pressure at which an 7TB"............. Fire controlled-small flickers only explosive mixture with air will occur in a confined 7'45"............. Fire nearly out--two small flickera vessel. 8'25"........ . Fire extinguished Ignition sources 8'56" ........ Foam stopped The test was considered successful on all counts. Other Sources of ignition should be eliminated from sty than the extinguishment of the fire, original areas of rene storage areas, no smoking rules adhered to, and concern included the possibility of undesirable reac totally enclosed electrical equipment used.-Explosion tions ot degradation of the monomeer from the foam proof equipment is recommended only in conAned and the extent of entrainment of monomer in the foam. areas. Static electricity is probably one of the more Neither of these proved to be of consequence. The sty important ignition hazards. All tanks, pumps, etc. rene was very simply reprocessed to a salable product should be adequately grounded and submerged filling and entrainment proved to be substantially less than practiced. that experienced with gasoline and hexane. Styrene tanks should, in general, conform to the Figure 7 illustrates & schematic of a typical storage Flammable Liquids Code NFPA-30. Tanks of over 10,OOO bbls. capacity should be individually diked, the dike having 100% of tank capacity. Dike drains are tank used by The Dow Chemical Co. Vinyl chloride .kept normally closed but with valve accessible under Vinyl chloride monomer (VCM) at normal ambient emergency conditions. Pumps should be located outside temperature and pressure is a colorless gas with a faint the dike. Distance between tanks should be a minimum sweet odor. VCM is stored and shipped as a liquid un of one half the sum of their diameters. der pressure and the greatest hazards in the handling From the standpoint of polymer buildup, the number of this material are the dangers of fire and explosion. of roof nozzles on a tank should be the minimum. A Table IV gives the more important properties of flame arrestor and vacuum-pressure relief are usually VCM as related to storage and handling. The same mounted on the seme nozzle. Emergency relief is com* three major objectives which apply for styrene mono bined with a manhoie('). A gauge hatch is the only mer also apply for developing technique and equipment other necessary opening, but a float operated tape level for safe handling and storage of vinyl chloride mono indicator is frequently provided. Flow line and drain mer: plus 3ome techniques peculiar to VCM due to its valves adjacent to the tank must be steel (preferred) or ductile iron as per API 604. high vapor pressure and extreme flammability. The objectives are: Fixed foam protection is the recommended provision 1. Personnel protection. for extinguishing styrene tank fires. The foam cham 2. Protection of quality and monomeric status. ber is separated from the tank by a frangible such as 3. Protection against fire and explosion. aluminum foil or glass to prevent polymer accumula Adequate training in the potential hazards of the tion. Inspection is at least annually. Provision should material and the proper personnel protection equip be made to supply liquid foam solution at a rate of 0.1 ment and techniques for it are essential to the safety gal./min./sq. ft. of surface. of all personnel handling VCM. The vapors should be In September 1967, Dow, in conjunction with Na controlled to no more than 500 ppm for repeated 7-8 tional Foam Systems Inc*' conducted a test to deter hr. daily exposures. mine the feasibility of subsurface foam injection for Concentration above this figure can cause dizziness, extinguishing fires in tanks containing styrene. Essen disorientation and "drunkenness." Concentrations ap tially, the method consists of pumping a foam solution into the tank through a product line. As the foam en proaching the lower explosive limit (3.6 vol. % in air) can cause helplessness and unconsciousness from a ters the tank, it rises to the liquid surface, providing very short exposure (anesthesia). a fire smothering blanket. It has generally been ac Thus, opermtimr areas must be designed with ade knowledged that such a system, if effective, would offer quate drainage and ventilation and equipment should cost, maintenance and reliability advantages over the be selected to minimize the possibility of leaks and commonly used foam chamber systems. These require spills. Operating personnel should be equipped with a separate, external piping system arranged to dis* charge foam on the liquid surface inside the tank. Ad and trained in the proper use of respiratory equip ment. mitting the foam through a swing line, placing it near Organic vapor canisters have been shown tD be on- as AP00022004 dependable and ineffective against vinyl chloride va pors. Self-contained breathing apparatus supplied with air or oxygen with full face piece and approved by the U. S. Bureau of Mines for this purpose should be used for protection against vinyl chloride. Liquid VCM spilled on the skin can cause frostbite due to rapid evaporation. Moderate chemical burns can occur to the akia due to liquid exposure, particularly with inhibited VCM, since the most commonly used in hibitor is phenol. Skin areas splashed with VCM should be immedi ately washed with soap and water; eye exposures must be Immediately washed with water for 15 min. mini mum. and medical attention obtained immediately. Contaminated clothing, gloves, boots, etc. should be removed immediately to avoid prolonged skin contact. Trained operating personnel must be fully aware of the flammable and explosive characteristics of VCM to preclude the possibility of human error causing a ma jor spill, or providing an ignition source for minor leaks or spills. Safety considerations Summing up VCM storage systems, should be designed to minimize leaks and spills; have adequate ventilation and drainage to conduct any leaks or spills away from critical operating areas; and the storage and process areas treated strictly aj hazardous, both from an engineering and operating procedural stand- print. AS personnel involved should be: well trained in the nature and hazards of the monomer and in the proper handling procedures; equipped with suitable eye pro tection such as safety glasses with side shields or their Equivalent; have adequate respiratory equipment-for emergency use; and provided with ready access to an eye wash and safety shower. Vinyl chloride is marketed under the specification shown in Tabla 0. For comparison, a typical analysis as manufactured la included. Unto 1956. essentially all VCM was inhibited as pro duced with 50-100 ppm of phenol to preclude the pos sibility ofspontaneous polymerization during handling and storage. During the late 1950's it became apparent to most producers that anihibited VCM could be stored and handled safely under the proper conditions. Since deleting the inhibitor makes the inhibitor re moval system at the users plant site unneessary, a considerable capital and operating expense Incentive exists to do so. Thus, a major proportion of VCM is now manufactured, stored and shipped in the uninhib ited state, although some users still specify phenol im- hibted monomer. .. Since large volumes of VCM are now stored and han dled in the uninhibited state, the proper design, main tenance and operating procedures for storage systems to protect the quality and monomer status of the stored VCM are absolutely essential to safety and loss pre vention. ndk., - Coincidentally these same steps also optimize the conversion to and quality of the end use product. Like styrene, the only significant end use for VCM is as a polymer iPVC) or co-polymer with other mono mers. The impurities or monomer characteristics which affect the polymer quality adversely are mainly color, oxygen, honupnfymer in the monomer iron content, and unsaturuted hydrocarbon impurities which produce figure 6. Conditions for explosive mixture with sty rene under air. variable polymerization rates and polymer quality. High water content is also undesirable. These characteristics can all be controlled within specification limits by the proper engineering and operating control of the final stages of the VCM finish ing train and storage facilities. Construction materials of the system can all be car bon steel. Indeed all equipment, instruments, guages etc. must be scrutinized to exetude the use of copper and copper bearing alloys due to the possibility of trace acetylene reacting with copper to form copper acetylides. Aluminum and aluminum alloys must also be ex cluded due to its reactivity with VCM. AU, valves, direct connecting instrument cases, pumps, casings, etc. must be carbon steel or equivalent, and not cast iron or ductile iron, to preclude the possi bility of major spills from equipment fracture or breakage. Stainless steels are acceptable but unncessarily cost ly for this service due to the size of most systems and in view of the suitability of carbon steels.' It is men tioned only because some minor specialty items such as Instruments may be more available in stainless steel. Storage tank pressure For ambient condition storage systems, design pres' sure of the storage tanks should be 100 psig, with 150 psig flanges, fittings, etc. used throughout the system. Horizontal cylindrical tanks are generally usfd for relatively small installations, 50,000 gal. or less. For larger installations, up to multimillion Iba^ spheres are preferred. In the case of vapor rccompression and refrigerated systems, lower design pressures can be used for the storage tanks, but pumping and piping systems should still adhere to the 150 ib. design. For safety reasons as well as monomer quality pres ervation, the oxygen content of stored vinyl chloride vapor phase must be maintained below 1000 ppm by volume. Thus storage tanks must be cleaned, dried and inerted to remove oxygen prior to introducing VCM into the system. Experience has indicated that the best way to remove atmospheric oxygen from a system is 57 AP00022005 NOTE- ROOF AND SIDE WALLS COATEO WITH AIR CURED EPOXY. PRODUCT OUT VACUUM PRESSURE RELIEF W/FLAME ARRESTOR GAUGE HATCH THERMO- WELL***., FOAM -- CHAMBER MANHOLE AND EMERGENCY ELIEF -TAPE-LEVEL INDICATOR 1 s THERMO WELL'! q---------- *-%. manhole swing JOINT-3&w. PRODUCT IN PACKAGE Q REFRIGERATION UNIT TO SLUDGE SEWER 1 H-- 1/2 TANK OIAMETER DIKE U-* 1/2 TANK DIAMETER Figure 7. Rec ommended de sign for styrene storage tanks. to pad and depad the system, (including tanka, lines, pumps, etc, as a unit) with nitrogen until the oxygen content reaches the desired level For example, padding a system originally containing one atmosphere of air to SO psig with nitrogen, depadding to atmosphere and repeating until the pad-depad procedure has been done five times will reduce the oxygen content of the system to 0.082 volume %. Experience has indicated strongly that vinyl chlor ide which is produced on the alkaline side, 0.2-0.5 ppm alkalinity as NaOH, has a more stable storage life. It is recommended that VCM which is to be stored or handled for any period more than a few hr*, be proc essed on the alkaline side. This can be achieved by pumping to storage via Hake NaOH traps. It is also recommended that flake NaOH traps be in stalled In the tank farm with the necessary piping systems to load builk VCM shipments via these NaOH beds. It is essential to protect VCM against contami nation by air, water, or any oxidizing chemicals, (per oxides or peroxide precursors in particular), to pre clude degrading the monomer quality. Steps to achieve this are obvious, i.e., do not permit tie-ins of any Of these deleterious compounds to VCM systems. VCM should not be exposed to sunlight, and storage temperatures should be maintained as low as possible. Sight glasses if used, should bs the suitably rated re flex type with actinic proof shields. However, Dow recommends the use of non-glass devices for-leve] in dication. A DP cell located at the bottom of'the tank with steam traced vapor leg connected to the tank top Is quite satisfactory. Tanks should be painted white or any other heat reflective paint system used to protect the monomer quality. In tropIcaLagsl. subtropical lo cations it is recommended that vapor recompression systems be used to maintain storage tank temperatures low enough to protect the monomer quality. Fire protection Vinyl chloride storage systems should be located in a segregated area well separated from the major proc ess unit and engineered as a Class I Group D Division I area and treated procedurally as such. Any sections of the process, or any adjacent process which has a higher than average fire or explosion haz ard, such as reactors or furnaces should be located as far away from a VCM storage facility as practical. The VCM storage system should be designed to elim inate and/or minimize the .possibility of leaks and spills by paying close attention to design pressures, safety valve settings, selection of valves, instruments and design temperatures. VCM should not be allowed to "free fall" from top entries into tanks. Inlets should be into the bottom of tanks or, if top inlets are used, grounded dip pipes pro vided. In addition, all tanks, pipelines and auxiliary equipment such as pumps and compressors must be grounded to preclude Ignition of leaks due to static electricity. A major VCM storage facility should be equipped with its own lightning rod. All operating equipment In a process unit and/or storage system must be inerted to preclude oxygen contamination prior to commissioning, and the obvious steps taken to preclude the introduction of oxygen slugs during service must be routinely followed. Examples are in loading tank cars, tank trucks, barges, etc, the receiving container vents hack to the storage tank. Therefore, these receiving vessels must be inerted to remove oxygen, preferably by padding up and down as already outlined. It is also Important that loading containers be elec trically grounded during the loading operation. It is desirable where feasible to have shipping con tainers returned from customers with a "heel** of about 5 psig vinyl chloride vapor to preclude the poasir billty of atmospheric air leaking into the container in transit It Is a good precaution to analyze the heel of VCM vapor in returned shipping containers for oxygen content prior to reloading. Storage tanks of greater than 10,000 USG capacity should be diked inside a con crete dike with.s capacity to contain the entire con tents of the storage tank, or tanks, if more than one tank is involved. Also, the diked area must have an underground drain equipped with block valves. This drain should not lead into the main plant sewer system as the danger exists of flooding the entire plant, in cluding ignition source areas with flammable VCM-air mixtures. Since vinyl chloride is lighter than, and es sentially immiscible in water, it will float on top of flowing sewer. The VCM storage dike drain should discharge to a 88 AP00022006 Figure 8. Sche matic diagram of a typical vinyl chloride storage system. separate underground sewer preferably leading* a re mote disposal or recovery sump. The drain block valve should be accessible from outside the dike wall, and left normally in the closed position; checked each day and opened when necessary to drain casual water from the diked area. Pumps and other auxiliary equipment, level gaugeff, etc. associated with the storage tanks should be located outside the dike wall. Major storage installations, say greater than 50,000 TJSG, should be equipped with an open head water del uge invoked by heat actuated devices and/or manually Invoked remotely to provide emergency flushing in the event of major spills or Area. Water alone will not effectively extinguish or control large VCM fires due to its lighter than water character. Water may even spread the fire by floatation, but will serve to protect the tanks and contents from overheat ing while the fire Is being brought under control by chemical means. In the event of a VCM Are. a point to remember is that one of the products of combustion is HC1 gas. It is recommended that in areas where mobile Are lighting rigs equipped with foam or powder capa- Tabte 5. Specification for vinyl chloride monomer. Property Sales Specification Typical AnaL Acetylene, ppm Acidity as HC1, ppm Aldehyde, ppm Iron, ppm Non Volatile Matter (% by weight) Phenol, ppm Inhibited Uninhibited Sulfur (as S), ppm Water (% by weight) 2 maximum 5 maximum " S maximum 0.5 maximum 0.05 maximum 25 minimum 100 maximum 2 maximum 5 maximum 0.03 maximum <1 0.5 nil <0.5 < 75 ppm 25-75 nil <5 <100 bilities are not readily accessible, fixed foam facilities be installed adjacent to the storage system and piped permanently inside the dike at required points to con trol and extinguish fires at any point within the dike area. Figure 8 is a schematic of a VCM storage facil ity. Summing up The best way to provide safety in VCM storage and handling facilities is to engineer the installation to minimize spills and leaks; preclude any ignition source for minor leaks and spills that might occur; provide a drained, deluged, dike area; guard against contami nating the system with air or oxidizing chemicals and train people thoroughly In the proper safe procedures for handling the product. # Literature Cited 4. UU*. 0ilk*," rutran. ana w. *>, --rraetieu way CO siM Safety Enf. (September IB, lHt). P. G. Shelley graduated from the Univer sity of Oklahoma in petroleum engineer ing and received an M.S. degree from the same Institution. Following two years as an Instructor at Pennsylvania State University he spent five years In research with Phillips Petroleum before joining Oow. He has been associated with styrene manufacturing and tech, nology for more than 25 years and is presently production manager for sty rene at Dow's Texas Dlv. E. J. Sills is a graduate in chemical en gineering from the University of Toronto and has worked in the Production De partment of Dow Chemical of Canada in the Chlorinated Hydrocarbon production area since 1951. He 1$ currently super intendent of chlorinated hydrocarbon process development. 69 AP00022007 Discussion Q. My ffrst Question has to (So with the problems of solu- bility of water in vinyl chloride monomer at typical storage temperature ranges from 40 degrees tor say. 120 degrees F. Do we have any data on problems there! SHELLEY; Say, f think t have it here in a bulletin. I am not really a vinyl ehloridt man. I can't answer that--I'd be glad to look or let you have e look at this bulletin after we finish here, and perhaps we can find the answer. Q. Right. Is that bulletin available! Could that be sent out? SHELLEY: This is a Dow Technical Service and Develop* ment Bulletin, and l think it can be obtained by writing to the Plastics Department of Dow Chemical Co., Midland, Mich. Q. The second Question has to do with the presence of Trace HCI in vinyl chloride monomer with respect to corrosion in carbon steel storage vessels. SHELLEY: Well, I can't answer that question either. \ know that as a general thing, we try to produee vinyl chloride and store it end ship it slightly on the alkaline side. This is accomplished by simply pumping the materials through a bed of flake eaustie. And just how much HCI it takes to cause trouble. I guess I couldn't say. Q. Would you straighten out my fuzzy thinking concerning these oxygenated compounds, the saturated oxygen and the TBC? And the use of air padding for tanks? I'm not quit# dear. SHELLEY: Wall, the TBC of eoursa Is basically a polymeriza tion inhibitor, but the nature of the inhibition, the chemis try of tnt inhibition, requires oxygen in order for it to work. And fortunately TBC is also a veiy good anti-oxidant for many different hydrocarbons. So it serves a kind of dual purpose in styrene. It holds down the formation of alde hydes and peroxides. The oxygen activates the TBC end permits it to do its job as an inhibitor. Now in storing under air the concentration of oxygen, if if* left then indefinitely, would get perhaps up to may be 50 PPM in the liquid. Bur ordinarily, as produced, the styrene has practically no air In ft And in a normal tank operation where you're shipping out of it every day, why, even under sir the concentration of oxygen doesn't get up much over maybe.Id or IS PPM, Q. I notice that under ordinary condition* the vinyl chlo ride is a gas heavier than air. I am a little curious as to the logic for dyking the tankage. SILLS: Well, vinyl ehloride is stored as a liquid under its own vapor pressure, and if you have a targe spill it will self-refrigerate, a small percentage of it will ffash off to. the atmosphere and coot the bulk of th liquid tfowrrso it will remain a liquid, and If you then can cover it-with* foam, you could even recover a large percentage of the vinyl that i has been spilled. But also the^dyking area prevents the spill from spreading massively Into ether areas whare you may have other flammable storage end spreading a fire condition, _ Q. I wanted to confirm whether I understood you correctly that polymer buildup and the relief devices and flame ar restors was not really a problem. You just minimized the number. SHELLEY: res. It's not a problem if proper inspection is fol lowed. Now i think that in most any of the tanks that I've been discussing hero, either under air or under natural gas or nitrogen, that these devices shouM bo Inspected about every six months and cleaned out Now they probably won't be completely closed up but there will be soma poly mer formation, and theca things have got to be cleened up every once in a while. Its just s metier of looking after your business, so to speak. Q. Well, the other point of clarification was, did I understand you right that you Size your relief devices as though these are flammable liquids only, not subject to heats of poly merization? A. R. ALBRECHT, Dow Chemical Co.: I'd like to add to the comments on the last two questions. First on the matter of flame arrestors, I think it specified in the paper presented that tha flame arrestor* should be Inspected at least annually. In practice, we should conduct that inspec tion frequency based on our experience. I know in the case of some styrene tanks we have to Inspect quarterly, be cause of the short period of time required to sustain poly mer buildup. So you have to be e little careful on this. On the second question regarding the sizing of emergency re lief vents, the calculation Is not based solely on flammable liquid criteria. We size our emergency relief vent* based on 1) heat input due to fire end Z) heat input due to poly merization, which in this instance msy exceed heat Input due to fire. This is detailed in Paul Shelly's reference, "Prac tical Way to Size Safety Discs" found in his paper. Q. I noticed in your article that you made quite a point of extending the fill lines in a sphere where you don't have any .oxygen. Is there some particular 'reason for that! it isn't standard LPG practice. SILLS: Yes, this may be "gilding the lily" a bit in the case of vinyl chloride, but It protects against the Inadvertent in* elusion into the tank system of a slug of oxygen. For In stance, If you loaded t tank car which had inadvertently been left with an atmosphere of air in it, your static build up on your failing droplets would not then ignite it The seme thing applies to grounding of the loading sta tions end piping end equipment throughout the storage sys tem. Here again with an oxygen free Inside atmosphere you shouldn't require grounding but if you have a leak, a Mown gasket or an escaped atmosphere, the grounding will prevent Ignition due to static, at that leak source. Q. What do you think of underground storage of styrene or simiiar monomers? SHELLEY: Well, I think it's all right. 1 don't really think If* necessary, f mean, there's certainly nothing the matter with It We've hed styrene tanks covered with dirt and-- dirt end concrete for many years, but usually It results in external corrosion and on* thing and another; it Just isn't worthwhile. Q. You mentioned tenk car Incident where the monomer went to polymer. You didn't mention the tenk blowing. Would you speculate why it didn't blow end what caused the polymerization? SHELLEY: Well, this happened I'd say at least 10 or more years ago, to tank'ear that was way oft down In Mexico somewhere, so I guest we realty don't know too much about it, but we speculate that all the oxygen was consumed and the TBC just failed to function and the ma terial started to polymerize, end of course being en exother mic reaction, (as I recall, It was an insulated tank ear), the heat of reaction accelerated the polymerisation rate, and finally it polymerized out of control. Now of course the more that it polymerizes, the thicker It gets and the more difficult it Is for the vapors to get away, and actually the greatest danger period is. say, whan there's maybe 30 to 60% polymer in the stuff. It just so hap pened that the relief valves on the tenk ear were adequate to take care of it. I don't know how close It came to blowing up the tank car, in other words. ALBRECHT: I want to add to the comments just made. 90 AP00022008 There ere probably several paperweights or solid polymer tanks scattered throughout this country. I'm speaking now e# buried styrene storage tanks--I don't know of any above* ground. I remember on# buried tank where there wes a polymerisation which whistled vapors through the vent line for over an hour. This is an absndoned solid polymer tank today. Q. We have tended to go away from flame arrestors on sty* rene tanks recently because of the possibility of the plug* fine of the vent t just wonder whether anybody else is tftmWng along the same firtesT We do use a pressure vac* uum relief, but not a flame arrestor. I reeliie there is one where the flame arrestor elements are outsida of the pres* sura vacuum relief element; but nevertheless we've tended to go to Just the one device. ALBRECHT: We are using flame arrestors just about uni* wndiy on our styrene storage tanks. I think there are many instances where perhaps we could safely do away with dame arrestors end use pressure vacuum conservation vent* instead. Q. Do you feel that foam system is justified in all eases for a small volume user? SHELLEY: Well, i would say it would depend on how much stuff you had around. I mean, we keep lots of styrene around without fixed foam protection. Now I think that around any kind of hydrocarbon plant you've probably got foam nubble that can be brought In, and I think that a find system probably is not necessary. I think that the tank shook* be sprinkled to keep it cool ip ease you get a fire - outside the tank. Okie ttfng which t didn't mention in the paper which I think Is of some significance is the distances we like to get vinjf cWorfde tanks awsy from things. Now relatively small stpm tanks, up to 20,000 galtons or so, we don't mind tavfng within, say, 50 feet of structures or operating areas. For larger tanks, we set minimum of somewhere in the order of 100 feet awey from any kind of e structure or op erating area. Vinyl chloridt we like to get even further away. In fact, in our big vinyl chloride plant that is partially un der construction, the main storage is about 1,300 feet away from the operating area. This Is a matter, of course, that you have to approach from a kind of a practical standpoint but i think good rule of thumb is the further away'you can get it the better. Q. What has been the fire experience with these monomers? SHELLEY: We've never had a styrene fire in the Texas Di vision. Now I can't speak far the rest of the company, but we've never had a styrene tank fire. We've had some firas in the operation resulting from spitls and that sort of thing, but nothing--nothing of any consequence. SILLS: We have never had a serious fire In the vinyl chlo ride installation to my knowledge. We have had minor firas at pump seals and the like, and these have ell been extin guished fairly readily, using a combination of CO. end dry powder extinguishers. And vinyl gives you quite a fierce fire, from even a relatively small leak source. However, we heve in recent years In Cansda had both styrene tank ears end vinyl chloride tank cars go through major plant explosion and in the case of the vinyl chloride ear It waa overturned and its dome wee buried in the ground, end they were under fire conditions for IS to 24 hours, and both the styrene tank ears and the vinyl chlo* ride cars were recovered'without any loss of product. They didn't go off specification, which was a real, I think, not commendation but encouragement to our method of han* dling in tank ears. They were in insulated' tank cars which protected them from fire damage. ALBRECHT: First water will not touch a styrene fire--you have to use foem to extinguish. Second, and this Is covered in the paper, we hive made some tests on subsurface foam Injection on styrene storage tanks and wa are going this route today In lieu of using the conventional fixed pipe foam system. *^-**MI 01 AP00022009