Document KJBDNEaNNDr6qNm9Qg5Q5gZ9X

Safe Storage of Dilute Ethylene Oxide Mixtures in Water James S. Curtis Hoechst Celanese Chemical Group, P.O. Box 58190, Houston, TX 77258 In the past industry has had to rely on a number of "rules of thumb" to provide a means of reasonable analysisfor many design, operation, safety and similar issues. The lack of readily available computingfacilities made the use of "rules of thumb" a necessary part ofdoing business in the hydrocarbon processing industry. And these "rules of thumb" have proven useful and allowed many tasks to be accomplished successfully. Today, however, availability of mainframe computers or desktop computing facilities has provided a tool that has decreased the need to rely on "rules of thumb. " And in many cases processing improvements can only be accomplished by replacing "rules of thumb" with more rigorous analysis. In the ethylene oxide producing/consuming industry a widely-used "rule of thumb" concerns the storage of ethylene oxide water solutions. A wellpublicized "rule" stated that solutions in excess of 1-2 weight percent ethylene oxide in water should not be stored [I]. This "rule" is certainly valid in many situations. However, this guideline may be unnecessarily restrictive in other cases. The impact of several key process parameters impacts the amount of ethylene oxide in water than can be safely stored. This study will show that higher concentrations of ethylene oxide in water can be safely stored under the right circumstances. The study will also discuss the key variables that determine whether aqueous ethylene oxide solutions can be safely stored. Finally, a methodology used in determining how much ethylene oxide in water can be safely stored will be outlined. FACTORS AFFECTING THE SAFE STORAGE OF AQUEOUS ETHYLENE OXIDE The factors affecting the safe storage of aqueous ethylene oxide solutions should be examined. First, the safe stor age of pure ethylene oxide will be reviewed. Ethylene oxide consumers and producers are generally familiar with the need to maintain an inert gas pad over ethylene oxide in storage. Ethylene oxide has a broad flammability range and even in the absence of oxygen, ethylene oxide vapor concentrations of greater than 75% in nitrogen form a flammable mixture [2], Thus, a safe storage temperature and inert blanket gas pressure are determined by the need to maintain the resulting ethylene oxide vapor com position below 75%, Figure 1 depicts the relationship betxveen the minimum blanket gas pressure requirement and a given storage temperature. A safety margin is in cluded in the data, but the intent of the relationship is to maintain ethylene oxide vapor compositions below 75%. Storing aqueous ethylene oxide solutions safely also re quires that the s apor composition of ethylene oxide be maintained below the 75% flammable limit when stored under a nitrogen pad. But, the "rule of thumb" dictating that 2 weight percent ethylene oxide in water solutions not be stored is not based on a vapor composition limit. The vapor composition of a 2 weight percent ethylene oxide aqueous solution at normal storage temperatures and pressures is much less than 75%. Yet the 75% vapor composition limit should be kept in mind as other factors influencing safe storage are examined. An aqueous solution of ethylene oxide will undergo a hydrolysis reaction. The reaction would proceed slowly Plant/Operations Progress (Vol. 9, No. 2) in the absence of a catalyst. But it will occur. As the exo thermic reaction progresses the temperature of the stored mixture will rise. A rise in temperature may then result in a corresponding increase in pressure over the liquid. A temperature rise and a pressure rise are two additional factors to be considered in analysis of safe storage for aqueous ethylene oxide solutions. Equipment integrity may be compromised by too high a temperature or, more likely, the inability to vent gas and prevent the overpres surization of a closed vessel becomes the critical concern. Again, the hydrolysis reaction would proceed slowly in the absence of catalyst. But, a "worst case analysis" would require that a contamination of the aqueous ethyl ene oxide solution has occurred that has promoted the Hy drolysis reaction. Still, the large excess of water would act as a heat sink and limit the temperature/pressure rise. The intent of this study is to outline the various process variables that impact the safe storage of aqueous ethylene oxide solutions. There are many possible reasons for storing ethylene oxide and water mixtures. There are processing steps in ethylene oxide manufacturing that result in the storage of aqueous solutions of ethylene oxide. Maintenance opera tions in ethylene oxide manufacturing and consuming in dustries may dictate a need for temporary storage of aqueous ethylene oxide solutions. Hoechst Celanese has utilized safe storage of aqueous ethylene oxide mixtures as a means of reducing environmental emissions gener ated when water washing process equipment in prepara tion for personnel entry and maintenance. There may be other reasons that dictate the need to store aqueous ethyl ene oxide mixtures. April, 1990 91 DO A 0 1 6 9 7 0 C O N F ID E N T !" Af R Note: Assumes insulated storage vessel. Figure 1. Storage prenure for liquid ethylene oxide under nitrogen blanketing gas. (Basis: <75% EtO in vapor space.) The storage of ethylene oxide-water material can be An analysis can be performed that determines how hazardous. This is the reason for the "2 percent rule." In much ethylene oxide in water can safely be stored that in presenting this analysis "worst case" scenarios will be as corporates the above process conditions as variables. The sumed. First, the storage vessel containing the aqueous analysis can be set up on an electronic spreadsheet, equa ethylene oxide may have a cooling capability for remov tion solver or other computer-based calculation method. ing the heat of the hydrolysis reaction and limiting the Then, case analysis can be performed to study the sensi temperature/pressure rise. This analysis will assume that tivity of each variable on the safe level of ethylene oxide the cooling capability is not available for whatever the in water for a given system. reason. Also, an ethylene oxide storage vessel will nor mally have the ability to vent gas to a vent gas scrubbing | facility. This analysis will assume that the vent scrubber is unavailable. The study will further assume that, for en- CASE ANALYSIS: TEN PERCENT ETHYLENE OXIDE IN WATER I vironmental reasons, the vent gas will not be vented to , the atmosphere. Instead, as the temperature rises the A calculational model was developed on an electronic storage vessel will increase in pressure because gas can spreadsheet to study the effect of the identified process not he vented. These worst case assumptions of no cool variables on the safe storage of ethylene oxide-water so ing capacity and a closed (no vent) storage vessel will pro lutions. Use of a spreadsheet allows one to alter a given vide the basis for the study. input variable, or set of input variables and "instantly" i Thus, three factors will be used to determine what is see the calculated results on the output variables. the safe level of ethylene oxide in water than can be The input variables for the model include: ' stored. First, an ethylene oxide vapor composition limit of Composition of mixture (weight percent EtO in water) ~5% must be avoided. Next, the storage vessel tempera Relative vapor/liquid volumes in storage vessel ture limits must not be exceeded should a hydrolysis re Initial temperature of storage vessel action occur. Finally, and most critical, the storage vessel Initial storage pressure of storage vessel pressure limits must not be exceeded. Fraction of ethylene oxide hydrolyzed to ethylene glycol Key assumptions used in the model are: KEY PROCESS VARIABLES AFFECTING THE ANALYSIS A hydrolysis reaction occurs (EO + HjO --* Ethylene Glycol) Certain key process conditions determine how much VLE of nitrogen-water-EO-ethylene glycol corrected r' ethylene oxide in water can be safely stored. These con for nonidealities gs l ditions include: Adiabatic reaction--no heat loss/gain to atmosphere i 1. The initial temperature of the stored material. Closed system--no venting of gas or dumping of liquid g - 2. The initial pressure of the storage vessel, No cooling capability exists L 3. The ratio of vapor volume to liquid volume, The output variables calculated by the model include: <X > 4. The concentration of ethylene oxide in water. The composition of the liquid and vapor phases _I 92 April, 1990 Plant/Operations Progress (Vol. 9, No. 2) 4----- -----1----- 1-----1-----1-----1-----1-----1-----1----- 1---- 1--I-----1------1----- I-----1-----1-----^ 5 Wt X EO in water 10 Wt X EO in water Basis of analysis: Initial vessel pressure of 244 kilopascals (before filling) Vessel pressure of 337 kilopascals (after filling, 10 X EO) vessel pressure of 311 kllopsscsls (after fining, S X EO) Initial vessel temperature of 30 degrees C i/i vapor to volume ratio Figure 2. Temperature rite in ttoroge vestel oi remit of hydrolysis reaction. The final temperature of the vessel The final pressure of the vessel Again, safe storage will be determined by examining the calculated output variables to see they do not exceed the following limits: Composition of ethylene oxide in vapor phase <75% Temperature below desired/design temperature of vessel Pressure below desired/design temperature of vessel Other important calculational methods built into the model include vapor/liquid equilibrium data for aqueous ethylene oxide solutions, hydrolysis heat of reaction data, heat capacity data, stoichiometry of the hydrolysis reac tion. ideal gas relationships for determining vapor density and pressure on the vessel, heat of vaporization data and other calculational steps. temperature rise from hydrolysis reaction First, the temperature rise resulting from a hydrolysis reaction was quantified. A 10 weight percent ethylene oxide in water solution stored in a vessel that had 1/1 vapor to liquid volume ratio, a starting pressure of 3 kilopascals (6 psig) and an initial temperature of 30C (86F) was used as the basis for determining the temperature rise. The results are shown graphically in Figure 2. The temperature rise shows a linear relationship as the hy drolysis reaction proceeds to completion. The maximum temperature reached occurs at 100% conversion of ethyl ene oxide. This temperature was 82.5C (181F). This temperature is below the vessel design temperature. Next, the concentration of ethylene oxide in the vapor as the hydrolysis reaction proceeds was calculated. Again, the basis was a 10 weight percent ethylene oxide in water solution, a 1/1 vapor to liquid volume ratio, an initial tem perature of 30C (86F) and an initial pressure of 3 kilo- Plont/Operotions Progress (Vol. 9, Ho. 2) pascals (6 psig). The results are shown in Figure 3. It is interesting to note the concentration "plateau" that oc curs as the hydrolysis reaction proceeds. The decrease in liquid ethylene oxide concentration and increase in vapor pressure caused by the temperature rise tend to balance out and result in a fairly steady ethylene oxide vapor com position. For this case the ethylene oxide vapor content was well below the 75% limit for non-explosive mixtures. Indeed, the ethylene oxide vapor content was not the lim iting constraint in determination of safe operation for any of the cases that the author studied. Either a high pres sure or temperature constraint occurs before the high eth ylene oxide vapor content limit does. Figure 4 demonstrates the effects the hydrolysis reac tion has on the closed vessel pressure. The storage vessel pressure rise as a result of the hydrolysis reaction is shown for 5 and 10 weight percent ethylene oxide in water solutions. Notice the plateau that occurs in the 70 to 80% hydrolysis region. The maximum pressure occurs before hydrolysis reaction completion because of the combination of ethylene oxide vapor pressure and solu tion temperature. Further cases will use 75% ethylene oxide reacted as the base "worst" case. The relative ratio of vapor to liquid volume was then studied. The relative amount of vapor volume available is critical as judged from Figure 5. Note that the analysis was made with 75% of the ethylene oxide reacting be cause the previous analysis showed that to be worst case in terms of maximum vessel pressure. As the relative vapor volume increases the vessel pressure rise de creases. Further analysis used a 1/1 vapor to liquid vol ume ratio. The effects of initial temperature are shown in Figure 6, Obviously, initial temperature has an effect on both maximum temperature and on the maximum pressure K O' 'C i l"t o achieved. Similarly, Figure 7 shows the effect initial ves <r sel pressure has on the maximum pressure achieved. Ini tial pressure is obviously an important variable in deter o c April, 1990 9.' flMP" T n r- h 5 Wt X EO in Water 10 Wt X EO in Water Basis of Analysis: Initial vessel pressure of 244 kilopascals (before filling) Vessel pressure of 337 kilopascals (after filling. 10 x EO) Vessel pressure of 311 kilopascals (after filling. 5 X EO) Initial vessel temperature of 30 degrees C 1/1 initial vapor to liquid volume ratio Figure 3. Ethylene oxide vopor composition resulting From hydrolysis re* action. 5 Wt X EO in Water 10 Wt X EO In Water Basis of analysis: Initial vessel pressure of 244 kilopascals (before filling) Vessel pressure of 337 kilopascals (after filling. 10 X EO) vessel pressure of 311 kilopascals (after filling, 5 X EO) Initial temperature of 30 degrees C 1/1 initial vapor to liquid volume ratio DO A 016973 Figure 4 Pressure rise in storage vessel os result of hydrolysis reaction. OONFIDFNTIAL 94 April, 1990 Plant/Operations Progress (V I. 9, No. 2) 5 Wt X EO In water 10 Wt X EO in water Fraction Of Storage veasel Used for vapor Basis of analysis: Initial vessel pressure of 244 kllopascals (before filling) Vessel pressure efter filling varies witn volume ratio Initial vessel temperature of 30 degrees C Assumed 75 X of EO reacted in hydrolysis reaction Figure 5. Storage veisel maximum pressure achieved as function of vapor to liquid volume ratio. *a Basis of analysis: 10 weight percent ethylene oxide in water Initial vessel pressure of 244 kllopascals 1/1 vapor to liquid volume ratio 75 X of ethylene oxide hydrolyzed (before filling) Figure 6. Effect of initial storage temperature on maximum vessel pres sure achieved. 00 A 01^974 CONFIDENTIAL Plont/Operotions Progress (Vol. 9, No. 2) April, 1990 95 Basis of analysis: 10 weight percent ethylene oxide In water Initial temperature of 30 degrees C 1/1 vapor to liquid volume ratio 75 X of ethylene oxide hydrolyzed Figure 7. Effect of initial vessel pressure on moximum pressure achieved from reaction. *a Basis of analysis: 96 April, 1990 Initial vessel pressure of 244 kllopascals Initial vessel temperature of 30 degrees C 1/1 vapor to liquid volume ratio 75 X of ethylene oxide hydrolyzed (before filling) Figure 8. Maximum vessel pressure achieved for varying levels of ethyl ene oxide in water. 00 A 016975 CONFIDENTIAL Plont/Operations Progress (Vol. 9, No. 2)