Document 44q7oea36do5a9DBV30nE47MV

iCicstd ".V>;f'"i--.'- 'Lz2 ; 93i SO 32 (*. 4-47)--fritifd in U.S.A. SHELL DEVELOPMENT COMPANY Kf(NCC i fo-c.* O'oO to SHELL CHEMICAL COMPANY SAN RAMON MANUFACTURING-PLANT & ANAL.PROJECT COORDINATOR-EL PASO DATE NOVEMBER 10, 1976 FROM DEPARTMENT HEAD-FOPMULATION/ COMMERCIAL FORMULATION SERVICEBSRC-MODESTO 1 Oi^ lVA *T SUBJECT MINUTES OF SAFETY DESIGN MEETING EL PASO PLANT; OCTOBER 20-21, SAN RAMON Per your request, your draft of the minutes of the October 20-21 meeting at San Ramon covering discussions of safety in grinding facilities at the El Paso Plant is returned with comments inserted -on the margins by the Modesto attendees. We are in agreement with the recommendations as set forth there including our margin inserts. Some additional comments follow. \ In reviewing the safety features of this plant, it becomes of primary importance for all areas of the company to appreciate both now and in the future that this facility is a special purpose grinding facility. It does not have the safety features for grinding materials that have a greater dust explosivity hazard than BLADEX Herbicide. We recommend close attention to the flammable vapor (DMK) content of the technical herbicide as introduced into the blending facility of the plant. Per the discussions we agree with the proposal for close surveillance of the incoming feed for its solvent content. As this monitoring apparently will be on a manual inspection basis, we raise the advisability of a flammable vapor alarm system for the units which preceed the inerted grinding loop. An unknown which could have significant impact on safety design parameters is the level of flammable volatiles given off in thermal decompo sition of BLADEX or atrazine. We are in the process of determining these and will keep you current with our findings. As you are aware, the BLADEX dust explosivity characteristics which are critical to the safety design ofthe plant have been determined necessar ily on material other'than Mobile production. It may be prudent to recheck the values of critical importance as soon as this material becomds available in order to provide confirmation of the design basis. We appreciate the participation in these discussions, and regret the delay in returning this memo. Attachment cc: Shell Chemical Co., San Ramon Information Services (2) T':3irdl Signed by G. W. GAEKTNER' G. W. GAERTNER LAM 027518 r<Jli . I PLAINTIFF'S EXHIBIT ABS-009773 San Ramon MINUTES OF MEETING DRAFT Shell Development Co. _______Modesto______ slOCT 5:71976 1S/0-7(r ^ October 20-21, 1976 Attendance - per attached list. Purpose: To review the design of the El Paso formulation plant and the Mobile triazine plant concerning safe handling of triazine dust and powder, especially with new explosivity and thermal decomposition data. To reconcile -a proposed BLADE! grinding safety guide and the above designs. The attached Agenda was followed. Results New explosivity data for more finely ground herbicides, i.e., BLADES and atrazine wettable powders--while not correlating well theoreti cally with earlier data--indicate that the explosivity factors are of the same order as for earlier data. The heat of decomposition of BLADEX appears to be about 25, OOqQa^C-C calories/gram mol exothermic, enough to raise the temperature 347 adiabatically, excluding melting. Differential thermal analysis indicates that up to 200C "other ingredients" do not affect BLADEX decomposition. Slow and mild exothermic decomposition occurs above or beyond 200C for W.P. and 240C for technical BLADEX. Atrazine. is in same general range. Little is known about decomposition gases--no HCN has been observed so far--more work is needed. Defiled reports will' follow; preliminary presentations are attached. ^ v LAM 027519 Residual acetone concentrations of above 0.02% w can lead to explosive mixtures above 100F. The current specification of 0.057. w ABS-00977 4 T 2 volatiles, as acetone in BIADEX should be revised. In practice product c runs under 0.01/i w,.act El Paso incoming feed--especially in tote bins-- should be monitored for acetone content or explosive atmosphere. Mobile will strive to avoid such. "Commercial incidents" - thermal decompositions in BIADEX formu \ lation are documented in the attached reports from Helena- Chemical and a Modesto Trip Report. A recent thermal decomposition at DEGUSSA's techni cal BIADEX manufacturing facilities occurred in product drying. In all instances there were no reports of poisonings, injury or significant damage to facilities. Causative explanations may be: a. Operating with a blocked hammer mill. b. A piece of metal in a hammer mill - tearing screens - overheating extending to product. c. Overheating in a product drying system - (DEGUSSA) possibly exacerbated by some impurity. * LAM 027520 ^ No explosion incidents have been reported to date. There are no confirmed incidents regarding atrazine, only rumors of GEIGY decompositions. Mechanisms and sources of overheating or electric discharge were discussed; concluding the following: Herbicide powder, falljq^-djwm--en. a bin, is not likely to develop A. Tr*****- ^ a spark potential exceeding 100 milli-joules*/well belov; that need'ro ignite ^ Qf.tJ.IA~ dust clouds per attached data. Ground powder, flowing in grinding loops. s will gain heat from hammer-mil ling and may gain static charge flowing in the loopsj Ojxx l+Jsfoi ,fitUZL aldjJZr- ~ pjt piece of metal, coming loose in the system - or escaping de- tection, etc. - falling free in a large bin or rotating within a ribbo 1 blender will not heat or develop a spark intensity sufficient to cause ABS-009775 .. ckun^ - tf* ob~~ 1 I -3 ^ U'*W '/"*p. ?PF.- Tight seals on rotating shafts as in/star valves, feeders. screw conveyors and ribbon blenders are sources of heating. However, the y**Ajrv' Y'-t-'j <L4Jo~ul 4fl *. J& C^v^c-*. heat is removed considerably by the flowing powder and metal mass'. In a A ribbon blender the charge moves over the hot-spot drawing off the heat. (Some ribbon blenders will have water jackets.) Blender H-1028 at El Paso S may hold from 5,500 to 13,750 lbs. of wettable powder feed - has a 75 HP \ drive - if all energy put-in converts to heat 191,000 BTU/hr. and is held . " N,___ adiabatically,- the theoretical temperature rise is (ll5 to 46F/hr. J Product -- ' melts at 325F.and decomposition starts at 392 F. There- will be heat loss \ 0\i. t and melting will absorb heat. A blocked-in ribbon blender with or without tight shaft seals will heat up slowly. With extended mixing *it needs watching for possible overheating. The tendency of triazine powders to develop and to bleed off static charges is not known quantitatively or relative to other materials. Some measurements should be attempted by Shell Development to orient on this'aspect. Electric grounding in bag.houses, bins, lines, ducts, etc. is being provided and will be scrutinized for adequacy of electrical connections. The El Paso design for powdered formulated product manufacture ' and handling and for building air/vacuum clean-up systems was reviewed in detail and was accepted as satisfactory subject to some items below. By way of quick review, the design calls for: 1. Nitrogen tempering (reduced oxygen) of the gas circulation in each of the three hammer-mill grinding loops (includes oxygen analyzers and alarms). LAM 027521 ABS-009776 A -4- 2. No nitrogen in receiving or weigh bins, in feed- preparation ribbon blenders, product blenders "sack-off bin" or bagged feed slitting. 3. Electric grounding throughout. 4. Dust proof electrical systems, NEKA 12. 5. Magnets at strategic points with full metal detection and feed cut-offs at hammer mills. 6. Indoor bag houses with electrically grounded bags. 7. No rupture discs with large vent headers. LAM 027522 8. Fire hoses and extinguishers at strategic points. Items to be added and/or pursued are: A. Make certain ducts are electrically bonded. B. Put hose connections and manual shut-off valves on all v major''idust'^ equipment'or where a significant amount of dust can occur. Provide hose reels conveniently. ^.Provide self-contained air masks to permit coping with fumes from a decomposition or overheatingfH- <Sax/ C. Add gas purge on seals where erosion by powder is probable. D. Check unit heater motors for adequacy in dusty atmosphere. Correct any problem - use fresh air blanketing if need be so. E. Check sandmill "drip-proof motors" for compliance with dust- proof criteria. Correct if problem exists. F. Set up inspection incoming totes and grinding loops for presence of acetoneT^I^^*^*^^- b * *T) G. Provide temperature alarms in gas leaving heat exchangers in grinding loops to warn of poor or no cooling. _.h ABS-009777 -5- H. Provide thermometers on hammer mills to catch any local heat buildup for whatever reason. I. Provide for grinding loop automatic shutdown when the secondary filter has a high pressure drop. J. Provide grinding loop pressure relief -.protecting against over-pressuring by blockage, excessive nitrogen, etc'. K. Avoid skin temperature in building air (unit) heaters which melt or would decompose the triazines. L dLc^jyt 01 aJlo-K-N -+ <L, Mobile and Vicksburg were also reviewed with limited information on the latter. No difficulties arose re Mobile. Vicksburg should be advised of our safety concepts. The proposed safety guide and an amended version, more applicable to BlADEXg and to atrazine, were reviewed. It was concluded that the pro posed guide should be called a "general purpose pesticide grinding guide" as it was intended and was used originally for materials more sensitive than the triazines and by formulators considered much less skilled and diligent than the Shell plants. Copies of the relevant passages under consideration are attached. , general guide calls for: The differences are highlighted in that the I 1. Nitrogen tempering of air whenever BLADEX or atrazine is processed instead of only in grinding loops. LAM 027523 2. Outdoor location of bag houses - (suitable in warm southern climates). 3. Water addition by spray headers with remote activation versus plain hose connections. 4. Large blow-out rupture discs with vent-off headers. ABS-009778 submitted to Safety and Health before Operations & Distribution use it in obtaining custom formulation work. I Toxicology reported that acute inhalation studies with ground BIADD(t> showed no untoward effects so far - all test animals had not been examined. Measures to keep air clean down to 10 mg. and.better yet 5 mg/ cubic meter should not be relaxed. Noise control is not overlooked but difficult to tackle till the specific arrangements have been constructed. \ LAM 027524 ABS-009779