Document EaKRN743Z3Mbj1211Dxx2Z0x

August 29, 1956 Distribution! D. B. Hosmsr ^_T1. J. E. Crouch ns copy For-->Jack Clegorn Harold Hubbard -St. Louis Main Office - 2 copies Extra - 3 ANALYTICAL LABORATORY INVESTIGATION # li AROOLOR 12k2 - AFTER-CORROSION CHLORIDES y^POse Determine cause of differences between Anniston, WGK, and General Electric laboratories in estimation of After-Corrosion Chlorides. Method 1. Dlreot contact was made by telephone with Hudson Falls laboratory to obtain exact procedure followed in corrosion test and determination of ohlorides. Procedure of WGK laboratory was observed at time of visit In June. Further details were established by telephone. 2. Complaint samples were returned to Anniston. They were checked by both Anniston and GE procedures for corrosion stability and afteroorrosion chlorides. TOWOLDMONOQ57518 Test Sample (Original Cl") Lo(t .510) 1 Lo(t .510) 2 Lot 508 ( <:i) (Lot 5(1.21) Lot 506 ( <a) Lo(t 5<1.0l) Lo(t 5<1.5l) Lot 516 (.^.1 QE Teat) Lot 505 ( c*1) SUMMARY AFTER CORROSION CHLORIDES AR0CL0R 1242 Corrosion Treatment C??!!e Test 500 flask Anniston 250 flask GE General Electric Testing WGK Testing 500 flask (Anniston (GE 250 flask (Anniston (GE General Electric Testing WGK Testing 250 flask (Anniston (GE General Electric Testing WGk Testing 250 flask GE General Electric Testing 250 flask ' . GB General Electric Testing General Electric Testing General Electric Testing General Electric Testing General Electric Testing WGK Testing 8-27-56 2 ppm Cl" <4 <4 0.1 0.3,0.3, <.l 0.4 <4, <a <4Ti 012, 0.2 0.2, 0.2, 0.3 0.2, 0.2 0<.12, 0.2, C1 0.1 0.3, 00..34 0.3 .1 <1, 0.2, 0.3 (F.E4(H.F.)(Pitt. 0.3 0.3 <4, *4 ^.1 1 0356136 TOWOLDMONOQ57519 Analytical Laboratory Investigation #4 3 Summary 1. The following summarizes the procedures followed and the magnitude of after-corrosion chlorides obtained: Corrosion Test GE - WOK Method 250 cc flask 200 cc sample (300 g) 210C for 6 hours A-C Chlorides Found__ Chloride Determination GE - WGK Method equal weights Aroclor and water double extraction with water wash water extract with ether develop turbidity with silver observe Tyndall beam Anniston Method (See Note (1) 500 cc flask 300 grams 210C for 6 hours Anniston Method 200 g Aroclor - SO cc water single hot extraction with water wash water extract with ether develop turbidity with silver observe against standards with transmitted light Note (1) Anniston corrosion method is identical to the General Electric 1945 procedure. The change in flask size was made in 1947 without the knowledge of the Anniston laboratory. Note (2) Low chlorides (0.1 ppm or below) were obtained with all combinations of corrosion test and chloride determination except that specifying the 2S0 cc reaction flask and the Tyndall beam detection. 2. Comparative results of Hudson Falls, Pittsfield, and WGK all using the same procedure gave data ranging from0*l to 0*4 on the same material. General Electric laboratories disagreed to the extent that Fort Edward accepted Lot 506 with ^0.1 while Hudson Falls and Pittsfield rejected it with 0.2 ppm and 0.3 ppm, respectively. 3 Work done in 1947 comparing the General Electric chloride determination with the Anniston method Indicated that they were in agreement. It also showed that the Anniston method was more consistent and free of error. Conclusions 1. To duplicate General Electric values in After-Corrosion Chlorides, it will be necessary to use a 250 cc reaction flask in the corrosion test and to follow the Tyndall beam detection method. 0356137 Analytical Laboratory Investigation #4 4 2. There is no explanation for the pronounced effect of the change in flask sizes upon the corrosion stability. One hypothesis might be the effect of free space in the larger flask in vaporizing chlorides from the hot Aroclor. 3. The Tyndall beam detection method is best used as a specification limit test at 0.1 ppm chlorides. At this point the beam is only faintly visible. At higher concentrations, the intensity of the beam is difficult to estimate, making comparisons with standards unreliable in these ranges. The method is greatly affected by dust, lint, minute air bubbles, finger prints on the test tube, etc. The beam intensity Increases rapidly with time, requiring close time control, especially at the 0.1 ppm point where the beam is almost invisible. 4* For measurement of chlorides formed by thermal decomposition, an entirely different procedure should be developed. It should retain all chlorides formed and measure them by a more reliable technique. W. B. Dunlap 0356138