Document yYL5VYExngONKvNpYbqqoyeD

JUL Mb JUL 06 '95 IN. PE, 2/12 Glycol Ethers and Thermal Runaway GowlKt Reaction: RCH2CH2OH + BiseCN&OHorKOH)-------- KO^OIO + H* CO+CGfc+7? Actual IndteBi L TgtrahvdfpfurfbrH Alcnhnl Still IhddfintRgflifltlfla On November 2,1961, an explosion occurred at tbe South Charleston Plant during the final stages ofa distillation of tetnhjdroAuhuyf alcohol from sodium hydroxide. Caustic concentration (NaOH) was eight to nineteen percent In five minutes, the kettle temperature Increased S0*C (252C to 300C); the pressure increased from 13 mreHg to greater than 35 psig. The steam had been shutoff and the system was under nitrogen. ApparrnLCauge A caustic-ether decomposition reaction In a batch process. 2. Ethontrigivcol Still JncidcnLPcftriptifln On March 18,1964, an explosion occurred after mhaxytrlglycol recovery was completed. Steam to the calandria was shutoff. The vacuum had been broken with hydrogen-methane gas (6 mm Hg to 218 psia) and transfer of kettle residues to storage had begun. The kettle temperature exceeded 200C within 15 minutes after pressurizing the column. An Initial, small "bump" occurred. Safety valves set at 30 psig blew. Within a few seconds a larger explosion occurred. The Incident report claims air was drawn into the column between the two explosions. The kettle residue represented 0.7% of the original charge and contained at least 40% base (as NaOH). JL*_ Kto PO;5^l-n HH-t.lt.U ORGflNirc: JUL 06 '95 as-'ism oou oxide & der flbiiN, P.P.3^12 2- Appaiwt Cause A caustic-ether decomposition reaction In a batch process. JtO.VjiepTj Incident Pweriodnn Laboratory experiments, designed to explore tbe hazards of ethaxytriglycol phot distillations employed synthetically prepared mixtures ofcrude ethoxytriglyeol <S25% CAMn,OL/SLS%ethcxytriglycol and 13.0% ethozytetraglycol) containing one to three percent sodium hydroxide. These experiments shoved the decomposition appeals to initiate at kettle temperatures of 195*C to 200C. The final caustic residue content in these laboratory distillations range from six to 1L5% sodium hydroxide. In another distillation study of crude ethoxytriglyeol, a spontaneous Increase hi kettle temperature was observed at 195C. Calculations indicate the base level as sodium hydroxide was 3.1% at this time, 4. UCAft Solvt 1LM fMMhvI DTPROPAgMl Mil Incident PMgtotim On February 19,1973, an explosion occurred at the end ofe distillation from caustic residue Six consecutive batches had bean refined. Sufficient heavies had accumulated In the kettle and a final "tails cut" before discarding these residues had been completed. One hour prior to the incident, the operator noted that the head pressure (10 mmHg to 130 mmHg) and the kettle temperature (120C to 195C) had increased during the preceding hour, lie operator turned off the ealandria steam. The explosion occurred an hour later. The caustic level was 22% to 26%, calculated as sodium hydroxide, and the soluble iron concentration was 0l31% (3100 ppm) at the time of tbe explosion, to the laboratory, one sample, obtained from the Still before the incident, was stripped of all volatile material up to 190C simulating the kettle contents at the time of the explosion and submitted for a thermal stability test This sampie underwent a strong exothermic reaction at 19S*C. In one minute, the temperature reached 426C and the pressure in the dosed system increased 400 ptig. DO A 017337 CONFIDFNTTAL JULS3MXN. P.P.4^12 hazards in processing alkaline CRUDE GLYCOL ETHER STREAMS 1. PROCESSING THESEALKALEVE STREAMS AT ELEVATED TEMPERATURE MAY RESULT IN UNCONTROLLED EXOTHERMIC RUNAWAY REACTION WITH CONCURRENT PRODUCTION OF NON-CONDENSABLE GAS. 2. THE RESULTING TEMPERATURE INCREASE AND/OR GAS PRODUCTION CAN RAPIDLY LEAD TO DANGEROUS OVERPRESSURE AND SUBSEQUENT RUPTURE OF PROCESSING EQUIPMENT IF ADEQUATE SAFETY RELIEF IS NOT PROVIDED. OCCAF BTK-1 DO A 017338 CONFIDENTIAL HAZARDS IN PROCESSING ALKALINE CRUDE GLYCOL ETHER STREAMS 3. THERMAL DECOMPOSITION STARTS WITH THE NET OXIDATION OF THE GLYCOL ETHER TO THE CORRESPONDING ALDEHYDE AND HYDROGEN. 4. COMPLEX SUBSEQUENT REACTIONS THEN OCCUR TO FORM ACIDS, OTHER ALDEHYDES AND CONDENSATION PRODUCTS, WHICH EVENTUALLY LEAD TO AUTOCATALYTIC FREE RADICAL DECOMPOSTION! UCCAF JuILeeb,ss& &z$pn'D8wT^lkc&(^cSrwiN. CEE ci TIONS ARC * ACCELERATED RATE CALORIMETRY. PFj6s 12 ONSET TEMPERATURE = THE TEMPERATURE AT WHICH A REACTION SOLUTION REACHES A 0.02C PER MINUTE EXOTHERMIC REACTION RATE. RO-/ROH MOLAR RATIO OF GLYCOL ETHER ANION TO FREE GLYCOL ETHER IN ALKALINE GLYCOL ETHER STREAMS. 0CC1P BTK-3 00 A 017340 CONFIDENTIAL JUL KJt> JUL 06 *95 P-P.7/12 IMPORTANT GENERALIZATIONS ABOUT THE EXOTHERMIC DECOMPOSITION OKALKALINE GLYCOL ETHER STREAMS 1. THE ONSET TEMPERATURE OF THE EXOTHERMIC DECOMPOSITION REACTION DECREASES AS THE ALKALINITY INCREASES. 2. THE ADIABATIC TEMPERATURE RISE FROM EXOTHERMIC DECOMPOSITION AS WELL AS THE RATE OF THE TEMPERATURE RISE INCREASES AS THE ALKALINITY INCREASES. 3. POTASSIUM ALKALINITY IN GLCYOL AND GLYCOL ETHER STREAMS RESULTS IN SIGNIFICANTLY LOWER ONSET TEMPERATURES THAN SODIUM ALKALINITY AT EQUIVALENT LEVELS. 4. SOME EVIDENCE SUGGESTS SOLUBLE IRON OR INSOLUBLE IRON AS RUST AT LEVELS OVER 1000 PPM MAY LOWER THE EXOTHERMIC DECOMPOSITION ONSET TEMPERATURE 10-15C. 5. THE POSSIBILITY OF CONCENTRATING IRON TO OVER 1000 PPM IS MUCH HIGHER IN BATCH SYSTEMS, PARTICULARLY AFTER EXPOSURE OF THE SYSTEM TO AIR AND/OR WATER DURING SHUTDOWNS AND TURNAROUNDS. UCC&P BTK-4 A 017341 confidfnttai. Jm_ no no wi-iurn rt-ruttu ukuhinil >-, -- JUL 06 '95 0221PM DOW ONIDE&1IN. t ......... IMPORTANT GENERALIZATIONS (CONTINUED') pfL.a/iz 6. THE STABILITY OF THE STREAMS IS SUBSTANTIALLY IMPROVED BY COMPLETE NEUTRALIZATION OF THE ALKALINITY WITH PHOSPHORIC ACID TO PH 1A AND COMPLETE REMOVAL OF THE INSOLUBLE PHOSPHATE SALTS. 7. THERMAL STABILITY TESTING HAS ALSO SHOWN THAT EXCESS PHOSPHORIC ACID SIGNIFICANTLY, ALTHOUGH TO A LESSER DEGREE, LOWERS THE THERMAL STABILITY OF THESE STREAMS, THUS CAREFUL NEUTRALIZATION IS REQUIRED. 8. STRONG ALKALINITY CAN BE REGENERATED IN SITU FROM NEUTRAL SALTS DURING A PROLONGED BATCH DISTILLATION. 9. AT LEAST TWO EXOTHERMS EXIST IN ALL CASES. THE TEMPERATURE RISE PRODUCED BY THE FIRST EXOTHERM CAN BE SUFFICIENT TO INITIATE THE SECOND MORE SEVERE EXOTHERM. 10. PURE GLYCOL ETHERS WILL EXHIBIT EXOTHERMIC THERMAL DECOMPOSITION WITH ONSET TEMPERATURES AS LOW AS 290C. UCCJLP BTK-5 DO A 01734,? CONFIDENT! Al WWW WW JUL 06 '95 PF.,9/12 BUTYL DIPROPASOL SOLVENT SODIUM ALKOXIDE DO A 017343 CONFIDENTIAL JUL Mb 'ab JUL 06 '95 * BUTOXYTRIGLYCOL POTASSIUM ALKOXIPE 12 1 i 1 I 1 i I i irt i i i i 120 140 160 180 200 220 240 260 ONSET TEMPERATURE (C} DO A 017344 CONFIDFNTTAL JUU (yo JTJL 06 '55 BUTOXYTRIGLYCOL SODIUM ALKOXIDE ONSET TEMPERATURE (Q