Document MJadBK4xNj3J6qyrnVp4EDO1z

B-2511 August 15,1995 Dow Confidential The Dow Chemical Company F 0 Box -OO PlaQuem-e Louisiana 7C7fcS-0-0Ci Report # LCPP95-016LR J.H. Albuquerque, Bldg. 5210, Aratu Richard Alvarado, B-2234, Freeport Bill Cook, A-1230, Freeport Rejeanne Cool, Bldg. 158 Ft Sask. Scott Daigle, 2601, Plaquemine Dean Grigsby, B-2512, Plaquemine John Gross, B-2511, Plaquemine Elaine Haun, 3601, Plaquemine Boyd Horton, B-2509, Plaquemine Gretchen LeBlanc, 3601, Plaquemine Herbert Leichert, Bldg C-5, Stade Leslie May, B-1463, Freeport Klaus Menschig, C4-3, Stade Roger Page, OC-201, Oyster Creek Larry Parrett, B-2224, Freeport Rick Thornhill, 2601, Plaquemine Cal Towns, Bldg Unit C, Ft. Sask. Ted Westlake, B-2511, Plaquemine John Young, A-1230, Freeport Nitrogen Trichloride, High Levels at CAII The Chlor-Alkali II plant in Louisiana had experienced higher than normal levels of nitrogen trichloride (NC13) (31,000 ppm) in the taffy pots after a complete dump of the reboiler and taffy pots in mid May 1995. In June of 1980 the D-227W taffy pot at CA II exploded due to an over pressuring as the result of a thermal run away reaction in which nitrogen trichloride played a significant role (1,4). As a result of this learning experience a more cautious approach was taken when it was noticed that days following the dump of the reboiler at CA n the measured value of nitrogen trichloride climbed above usual values of around 1000 - 5000 ppm. Following the complete dump of the system, spikes were noticed in the concentration as high as 16,000 to 31,000 ppm. In order to prevent a dangerous condition from occurring, DO 072913 CONFTDFNTTAL. the reboiler was initially dumped daily and later every 48 hours to keep the NCI3 from concentrating. With each dump of the reboiler a grab sample was taken of the taffy material. Each sample was analyzed for NCI3. A table of the concentrations is in Table 1. Sample Date Feb-03-95 Feb-16-95 Mar-02-95 Mar-16-95 Mar-30-95 Apr-13-95 May-12-95 May-13-95 May-16-95 May-17-95 May-18-95 May-19-95 May-20-95 May-21-95 May-23-95 May-25-95 May-27-95 May-28-95 May-30-95 Jun-01-95 Jun-05-95 Jun-13-95 Jun-16-95 Jun-21-95 Table 1 Nitrogen Trichloride in Taffy Pot Temperature in Reboiler. F 160 168 48 99 109 42 43 32 32 31 31 31 30 31 33 34 33 30 31 31 31 37 36 152 ppm Nitrogen Trichloride 784 772 525 1182 849 5185 6139 17045 31000 11700 16900 18450 17400 19500 16631 5526 30983 10100 21200 18723 19491 20200 27224 3343 Reboiler temperatures were in the 30 to 40F range which are well below the temperatures needed to decompose NCL3. The most effective decomposition of NCI3 occurs in the neighborhood of 150-170F. Noticeable temperature increases had not occurred in the reboiler and taffy pots since the pots were dumped frequently in the days following the concentration spikes. The previous time that a complete dump occurred of the reboiler and taffy pots was in May, 1994. It took 39 days for the reboiler to reach 170F. During the 39 days, NCI3 levels were not measured. Based upon the NCI3 levels seen after this latest dump, the plant personnel did not want to risk waiting this long time before the temperature would build-up. 00 0?293 4 CNFTDENT Tal To obtain a better understanding of the nitrogen trichloride excursion it was necessary to determine how the NC13 was forming. The nitrogen trichloride that was present was forming in the cells as a result of the nitrogen containing compounds in the incoming brine being chlorinated. A majority of these compounds are ammonia, earth borne acids, and decomposing organic compounds entering die brine wells. A possible route of formation in a chlorine containing aqueous solution is described in Equations 1-3. HOC1 + nh3 ---> NH2C1 + H20 Eqn. 1 HOC1 + NH2C1 ---> NHCL + h2o Eqn.2 HOC1 + NHCU ---> NC13 + h20 I*qn. 3 The reaction that predominates is dependent on the pH and the amount of chlorine present. With each successive reaction, a greater ratio of initial chlorine to ammonia and initial pH would be required, however, nitrogen trichloride is found to be formed below a pH of 5 with equimolar ratios of reactants only. Based upon a NC13 concentration in the taffy pot of 30,000 ppm and other variables specific to the Chlorine plant, the amount of available nitrogen in the incoming treated brine is estimated to be 0.23 ppm. In order to better justify the levels in the taffy pot, the incoming brine was analyzed for Total Kjeldahl Nitrogen (total available monatomic nitrogen). Analyses were performed by Westpayne Labs, Baton Rouge, LA. The results of the analyses are in Table 2, with the results verifying the levels of NC13 in the taffy could be coming from the nitrogen compounds in the brine. Sample Date 05-23-95 05-23-95 05-23-95 06-01-95 06-01-95 06-01-95 06-01-95 Table 2 Total Kjeldahl Nitrogen Sample Location LADC12 LADCAH LAD C12 & CA H LADC12 LADCAH LADC12&CAH LADC12&CAH Type of Sample Treated Brine Treated Brine Raw Brine Treated Brine Treated Brine Raw Brine Process Water ppm Nitrogen 0.28, 0.56 0.28, 0.70 1.50, 1.40 0.17 0.23 1.99 0.37 It became apparent based on the data presented that a solution to the problem was to raise the reboiler temperature to facilitate NCL decomposition. A recommendation was given to add a heavy organic such as carbon tetrachloride (CCL) to act as a heat sink to maintain the temperatures necessary in the reboiler to allow for the decomposition of NCI3. Carbon tetrachloride had been used before in this process, but was removed a year earlier in order n0 07291.5! confidentTM. \ to be in compliance with EPA regulations, the Montreal Protocol. However, since no CCL contacts the final product and the CCI4 is sent to solvents for disposal; CCI4 was allowed to be used in the reboiler as the result of a grandfather clause in effect for chlorine production. Dumping of the taffy pot continued on a less frequent basis (every 72 hours) until June 18, 1995, when carbon tetrachloride was added. Days prior to the addition of CCU the NCI3 concentration was in excess of 21,000 ppm in the taffy pot. Two days following the addition of carbon tetrachloride the reboiler temperature rose to 170 F. Temperature fluctuations continued until a steady state temperature of 160 C was reached. At that temperature nitrogen trichloride readily decomposes into nitrogen and chlorine gas. Nitrogen trichloride concentrations at this new temperature were measured at 3343 ppm. Lower concentrations are expected the longer the reboiler remains at this higher temperature. The effect of the addition carbon tetrachloride was significant in this application. The CCI4 had a three fold effect to act as a heat sink for raising the temperature in the reboiler, to dilute the reboiler's contents, and CCL4 has been shown to keep the heat exchanger free of fouling and prevent plugging of the line to the Solvents plant burner. References: 1. Gathers, J.J., and Horton, W.B., Investigation of Louisiana's Chlor-Alkali Taffy Pot Rupture - June 7,1980, LAD-LR 80.18,08/12/1980. 2. Gross, J.W., Meeting on High Levels of NCI3, May 19,1995. 3. Haun, E.M., Various Meetings on NC13 and Table 1 Data. 4. Hudson, T.B., and Castille, M.J., A Study of the reactivity of Chlorine Taffy Using Accelerating Rate Calorimetry, TC-875,11/12/80. as> Michael Mebes DO 072916 CONFIDENTIAL