ML20196J513
| ML20196J513 | |
| Person / Time | |
|---|---|
| Site: | Trojan File:Portland General Electric icon.png |
| Issue date: | 07/29/1997 |
| From: | SCIENTIFIC ECOLOGY GROUP, INC. |
| To: | |
| Shared Package | |
| ML20196J508 | List: |
| References | |
| CON-SFPDP-NQT81326-154, CON-SFPDP-NQT81326-154-0 SEG-TRJ-R-006, SEG-TRJ-R-006-R00, SEG-TRJ-R-6, SEG-TRJ-R-6-R, NUDOCS 9708040215 | |
| Download: ML20196J513 (88) | |
Text
- . . - - .. - - - . - - - _ _ _ . - - . - . -
_LM__M_
SEG/TRJ/R-006 Trojan SecondIntegrated Test Data Report l PORTLAND -
GENERA EL TRIC W RFf-bO3) 4.b go 5FPDP-NOf Rf AMo-l54-O teme no. SF*F0 9- 032.-97 L_
SUPPLIER DOCUMENT STAMP 1[
go inal submittel.
York may proceed Submit final docu-
.o One ano ,esubm,< wo,, me, ,,ocJHer MMk Rxashg Spam bi ect to incorporation of changes indi- e =
Oo kvise and resubmit. Work may not pro-eo % no, ,u,, .o,,me,p,ocee.. SpentFuelPoolDebris Project 6o ' )ther Perodosion to proceed does not co,sentote 9- - . ,__.;o or opproval of design deteRe, celo dations, onelyses, test methode, or me-
- lories 2;.- ;:I or selected by the supp-Ret and does not relieve suppiler from full
.: .a .m co e.t.ei onne s"*"*
Scientine Ecology Group,' Inc. !
D INFORMATION ONLY , . -
M COl11mhis Drive SE '
Prep f] JIPMand, WA 99352 Coni urred by NL * - '
Date 1 79 o,, , , , , , ,
5FPDP- NOI 813aG- 15 4 - O Project Project Rav. DCM No. _
Originador ,
Engbeer -- s,' @- r - Menayg Deee 0 -
khWY ,m lN $$
Ybf/ff V
TRJ arr-243 b0 3g 9700040215 970730 PDR ADOCK 05000344 Y PDR ,
-. - . ~. --- - - - - - . ~ . . -
_I. . bb_
TABLE OF CONTENTS Section Page 1.0 EXECUTIVE S UMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.1 B ac k g round . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 ,
1.2 Description and Results of Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1.3 Summary of Hydrogen Analysis Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 1.4 End-of-Run Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 2.0 INTEGRATED TESTS AT WNP-1 SITE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.1 Test Objectives and Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 2.2 Functional Run Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.3 Run #1 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.4 Run #2 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.5 Ru n #3 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 !
2.6 Run #4 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 2.7 Run #5 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 2.8 Run #6 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 l
2.9 Run #7 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 ,
" 2.10 Run #8 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 l 2.1 1 Run #9 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 8 2.12 Data Quality Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 i
3.0 RESIDUE TESTING AT WNP-1 FACILITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 L 3.1 Qua rtz Tube Apparatus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 l 3.2 Gas Chromatograph Calibration H2 .................................. 46 3.3 S ample Blan k . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 3.4 Trip Bla nk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 0 3.5 Hydrogen Quantitation Via Phenanthrene Sample . . . . . . . . . . . . . . . . . . . . . . 51 !
3.6 Run #1 Residue Biased Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 !
3.7 Run #1 Residue Random Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 I 3.8 Run #2A Residue Biased Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 3.9 Run #2A Residue Random Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 3.10 Run #2A Residue 2nd Biased Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 3.11 Run #3 Residue Biased Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 i 3.12 Run #3 Residue Random Sample . . . . . . . . . . . . . . . . . . . . . . . . ........... 57 3.13. Run #3 Residue White Coated Lumps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 3.14 Run #4 Residue Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 8 ,
i 3.15 Run #5 Residue Biased Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 !
- 3.16 Run #5 Residue Random Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 3.17 Run #6 Residue Biased Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
- 3.18 Run #7 Residue Biased Sample . . . . . . . . . . . . . . . . . .................... 61 l 3.19 Run #7 Residue "No White" Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 !
3.20 Run #8 Residue Biased Sample (cloth) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 TRJ RPT.2-03 Rev.O i
._ Page 2 of 87 i
i
SEG/TRJ/R-006, Troj n Secand Integrated Test Drta Repart L ,j @ @ @
--c u- - .c 3.21 Run #8 Residue Biased Duplicate Sample ("No Cloth") . . . . . . . . . . . . . . . . . . 64 3.22 Run #9 Residue Blased Sample of 1% of 0.25" PVC at 100*F Hotter . . . . . . . 64 3.23 Run #9 Residue Blended Random Sample of 1% of 0.25" PVC at 100*F Hotter.........................................................65 3.24 Residue Hydrogen Content Computation (Dross Correction) . . . . . . . . . . . . . 66 3.25 Data Precision, Accuracy, Completeness, Comparability, and Percent Relative '
Standard Deviation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 3.26 Thermogravimetric /GC Analyses at IT Lab . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
4.0 CONCLUSION
S AND RECOMMENDATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 4.1 Operational Comments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 4.2 In-line Hydrogen, Carbon Monoxide and Benzene in End-of-Run Determina tio'ns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 9 4.3 The Challenge of " Neoprene" (sic PVC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 5.0 A CRONYMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 2 6.0 REFE REN CES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 l APPENDIX A - DESCRIPTION OF TEST EQUIPMENT . . . . . . . . . . . . . . . . . . . . . . . . . 85 A.1 O verall Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 5 A.2 Component Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 i A.3 Process Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
(
l
? I l l l
e f
TRJ.RPT.2 03 Rn. 0 Page 3 of 87 1
.. -- - - . - -.-.~. .- - - . - . . , - - . . - -.- - -
, i i . l l SEG/TRJ/R-006, Trejen See nd Integrcted Test DIto R! port LM@@
x mo .
1.0 EXECUTIVE
SUMMARY
This test repon contains four sections: i i
l Section 1.0: Contains an executive summary of the Integrated Test results j Section 2.0: Provides key operational results from the testing conducted on the j full-scale, remote, process can handling equipment, the can feed '
evaporator (CFE) and the steam-reformer i
i
. Section 3.0: Summarizes the residue testing completed at the Integrated Test Site' ,
I ,.
Section 4.0: Provides conclusions and recommendations for operating at the~ j Trojan Nuclear Power Plant '
r 1.1 Background !
[ l The purpose of the Trojan project is to remove organics and other hydrogen bearing i compounds from wastes in the spent fuel pool located at Portland General Electric's (PGE) Trojan Nuclear Power Plant in Rainier, OR, and produce a concentratred I l residue with sufficiently low hydrogen content that any formation of radic4y*.2c ;
hydrogen in long-term storage casks is below 5 volume percent (vol %) of the contained void space. This is a Nuclear Regulatory Commission (NRC) requirement {
l1 '(IE 84-72 NRC,10 CFR 71) for long-term sarage. Note that this translates to about j l 250 milligrams H 2in a typical final capsule containing five process cans filled to a l
depth of 50%. "
The spent fuel pool storage racks contain debris with deteriorated cylindrical sock and i
, canridge filters loaded with clay sediment, activated metallic electron-discharge - l l . machining (EDM) material, and other organics contaminated with failed fuel element J fines and fuel pellets. The operations involve the remote handling procedures for the j recovery and segregation of this waste and the loading of steam-reforming process cans that are insened into the steam heated can feed evaporator. The waste is ,
processed to about 600*C (1125"F) in the can feed evaporator, the organic off-gases L are funher processed into syn-gas by the steam reformer (Galloway,1996).
There are four major steps in this project:
l e Treatability testing was completed in a small 6-inch tube steam-reformer system at International Technologies Corporation (IT, Corp.) Technology Development ,
Laboratory in Knoxville, TN. Each type of filter and flange provided by PGE !
was tested for treatability and potential operating conditions were assessed.
- e The Proof-of-Principle (POP) Test was completed at Scientific Ecology Group, L
Inc.'s (SEG) Bear Creek Facility in Oak Ridge, TN, in the commercial steam-l l
TRJ-RPT-2-03.
Page 4 of 87 1 e- - + v-- -r- -.
---7 m e 4
i l SEG/rRJ/R-006, Trojtn Sec nd Int: grated Test Dcto R:p:rt j@@@
_m, __
reformer to verify the ability of steam reforming to achieve the required low hydrogen content in the final residue of surrogate waste sock filters.
- The Integrated Tests were completed at the Washington Public Power Supply System (WPPSS) Nuclear Plant #1 (WNP-1) at Richland, WA, where the CFE and superheated steam-reformer (SSR) system were integrated and demonstrated with the full remote handling equipment to achieve the required low hydrogen content in the final residue.
e The actual processing of the Trojan Nuclear Power Plant debris.
1.2 Description and Results of Testing The Integrated Test runs were conducted May 12 - June 15,1997 in accordance with the Trojan Integrated Test Sampling and Analysis Plan (SEGffRJ/ PRO-032), the-Trojan SecondIntegrated Test Plan and Procedure (SEGITRJIPRO-033), and the Residue Hydrogen Analysis Protocol (SEGfrRJ/ PRO-006). A brief description of the steam reforming equipment and process flow for the test is provided in Appendix A.
A simplified drawing of the test system is provided in Figure 1.1.
s The POP testing identified the optimum processing temperature and flow rates. The !
Integrated Test consisted of nine runs designed to determine optimum operating pressures and reliable end-of-run indications. Hydrogen (H2 ), carbon monoxide (CO),
oxygen (O 2), and benzene gas analyzers provided in-line indications of steam reforming processing status.
A surrogate waste mix of polypropylene sock filters, EDM dross, and spent fuel pool water chemistry control salts was prepared in an identical manner for each run. The only waste items which varied from run to run was the type of stainless steel Raschig ring or Berl saddle (or none), and, at PGE's request, the amount and size of black filter flange material. The stainless steel packing was added to enhance heat transfer rates to shorten processing times. .
The black filter flange material originally identified as neoprene turned out to be polyvinyl chloride (PVC) with sulfur bearing crude oil as an additive. PGE requested that testing be performed to determine the steam reformer's ability to process the black filter flange material (referred to as PVC thoughout this document). The result was that 53.5 gram (g) of PVC may be added to a process can if it is cut up into 1/4 inch or smaller pieces. And, CFE operating temperatures must be raised 100* to 1125'F to ensure that the PVC is processed sufficiently to meet the residual hydrogen limit.
l The residue handling procedures used during the Integrated Test incorporated extensive precautions to avoid picking up residual room moisture or water j condensation on cold transfer bell or slide gate parts. This contaminant water l moisture could form radiolytic hydrogen. The process cans were removed from the TRJ-RPT 2-03 Rev.0 p
- - _ - - . - -- ~ - . - . - - - - . - - - - . . . - . _ . - - ~ . _ . . . . . -.
C i
SEG/TRJ/R-006, Trahn Secend Integrat:d Test D:ta Report j@@@
_t. __ _
CFE by the Transfer Bell and transferred into an inerted glove-box fer analysis of i residual hydrogen levels. The residue samples for hydrogen analysis were removed l and sampled to obtain both a uniform, representative sample and also a biased sample !
with preference to organics taken from the center areas of large lumps and white l coated lumps, if present, to deliberately bias the sample for possibly poorly or !
. incompletely processed waste surrogate. To assure against any room moisture !
contamination during the process can sampling, laboratory technicians worked l
through gloves in the glove box, unloading the process can within the slightly t pressurized argon-inerted atmosphere. l 2
l A handling blank consisted of inserting an empty quartz glass boat into the quartz ;
tube and measuring the apparent hydrogen evolution curve, and determined the l
" apparent" H2 background that should be subtracted from each sample's H 2 l
determination. Since this background level was small, for conservatism it was not i applied to any of the analytical results. Trip blank samples of anhydrous sodium, sulfate (Na2SO4 ) and MSA water-indicating sample tubes were used to attempt to i quantify the residual moisture pick up in the transfer bell, slide gates and during !
residue removal operations. Also, residue sample control tests were made with a l
known quantity'of phenanthrene to measure levels of H2 quantitation and recovery l
rates from the sample. l Table 1.1 summarizes the integrated test processing conditions that were varied from Run #1 through Run #9. The "End-of-Run" criteria were varied with each run. The important processing variables were steam flow through the process can and CFE steam chamber, and CFE gas temperature.
Table 1.2 provides the processing times of each of the process steps in the
- different test. The term " process time" is used to describe the chemical reaction time of the steam-reforming from " start run" to "end run." Note that the process times vary from 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> (hrs) and 55 minutes (min) to 58 hrs and 30 min. The total run time is the process time plus the cooldown time.
I I
l i
l -
i TRJ-RPT-2-03 Rev.0 l
, . . . , . . . _ ~ _ _ . . , . _ . . . . . - . ~ _ . . .
i Mobile Steam Reformer Process Flow Boiler Superheater Can Feed To Plant Evaporator Exhaust ; L
)k i 5 m
, r l
Nitrogen p E E j g Superheater v b. / 4 l 2.
h~
l
! ,. ,l Blowers tor inlet og .
.g Heat Exchangers From Reactor outlet --
Reactor FIGURE 1.1 Rev.O Page 7 of 87
SEGiTRJ/R-006, Troj:a Etegrated Test R; port TABLE 1.1 INTEGRATED TEST
SUMMARY
- PROCESSMG CONDITIONS TEST RUN s 1 2* 2A 3 4" 5 6 7 8 9 Hardware Pipe heaters Band Band Band None None Redent Redent Redlent Redent Redent Transfer bas heated no no no no no yes yes yes yee cold pwge CFE pipe dmin no no no no no yes yes yes yes yes Grappleleek repelred NA NA NA NA NA yes yes yes yes yes CFE peuge 4 4 4 4 4 4 4 4 8 8 CFE piugs w ortAce,lrt die none none none none none none 4,9f16* 4,3F none ncne Process Cen s 1 4 4 3 1 5 2 5 1 20 Process can sielus new new new new used used near used cleaned new Reachig rtnes yes yes yes yes yes none none none bed ooddes berleeddes Black fienge futer meterial, greme none none none 533.7 525.2 none 531.6 606.4 none 53.5 Heat up Method CFE & Pipe togelhor or CFE hatter yes yes yes yes yes no no no no no Pipe 100F heller then CFE no no no no no yes ps yes yes yes , ,
End of Run C:Gerte No. hrs <500 ppm H2 3 0 16.75 0.85 NA 4 4.25 34.6 4.25 14.4 No. hrs <200 ppm H2 2.8 0 0 0.13 NA 0,7 1 12.1 3 3.1 No. hrs O ppm H2 2.7 0 0 0 NA 0 0.1 5 0 0 Ave H2 for last hour -26.9 >4000 234 337 NA 173 SS 0 79.3 102.6 Ave CO for last hour 0 55.5 0 0 NA 0 0 0 0 0 '
Ave Benzene for last hour . NM NM NM NM NM NM 0.53 0.05 1.4 0.72 N2 Purge time, min 10 10 10 10 10 25 25 7 hre 80 25 Average Temperatures (ener heetup)
K109 Super heeler ouMot temperature 1056 1057 1058 1054 NA 1057 1055 1057 1061 1167 K101 Inlettocan 93 1000 943 947 NA 1C37 1034 1053 1034 1137 K102 CFEtemperature 1029 1067 1052 1046 NA 1053 917 1047 974 1128 t
K103 CFEouusttemperature 874 933 922 934 NA 871 824 896 793 950 Average flows
- FY 100 Totalsteemscfm 22.2 21.1 23.1 22.2 NA 20.9 21.3 24.2 13.2 18.1 i FY 101 Process Can ocfm 4.5 4.9 2.5 4.8 NA 2.4 5.3 1.2 3.3 4.1 FY 102 CFE Steam Chamber schn 17.7 16.2 20.5 17.4 NA 18.5 16.1 23 9.8 16.1 ,
- Faers melted whBe in etener for 21 hrs Residue chopped up & rerunin2A
" Feiers meted due to condensees, aborted i
NA: Not Appecable NM: Not Mesoured . t marr-2-03
% s d87 i
. . L
SEGfrRJ/R-006, Trojan Second Integrated Test Data Report TABLE 1.2
. RUN PROCESSING TIMES Sodium Functional
- RUN # Salt Run 1 2 2a 3 4 5 6 7 8 7A@ 9 4
2AN # 2 5 1 4 4 3 1 5 2 5 2 1 20 Can Loaded 5/12 5/14 5/16 5/18 5/20 5/21 5/24 5/31 6/4 1 6/5 6/8 6/9 6/15 2300 2145 1931- 1445' 1300 1325 1120 0418 1050 2050 1945 1842 1145 Start Run 5/14 5/15 5/17 5/19 5/20 5/21 5/24 5/31 6/4 6/5 6/8 6/9 6/15 1017 1300 1219 1152* 1425 1350 1141 0456 1106 2107 2352 1900 1210 Xic:are End Run 6/1 6/5 6/8 NA 6/10 6/16 0425 0000 0330 0315 1215 End Run 5/14 5/15 5/17 5/20 5/21 5/22 5/24 6/1 6/5 6/8 6/9 6/10 6/16 1615 1953 1914 0510 0700 0840 1525*" 0825 0405 0725 0930 0515 1422 End Cooldown . 5/14 - 5/15 5/17 5/20 5/21 5/22 5/24 6/1 6/5 6/8 6/9 6/10 6/16 1805 2200 2125 0738 0853 1046 1725 1036 0643 0917 1130 0825 1719 rransfer to 5/14 5/16 5/18 5/20 5/21 5/23 6/1 6/5 6/8 6/9 6/10 6/16
~
31ovebox 2045 1930 1530 1000" 1010 1110 1300 1025 1035 1505' 0950 1746 Jnload Debris 5/15 5/16 5/18 NA 5/1 5/23 6/1 6/5 6/8 6/9- 6/10 6/16 i 1300 2150 2300 1300 1150 1400 1115 1120 1530 1045 1930 3rocess Length 6 hrs 6 hrs 17 hrs . 16 hrs 18 hrs 27 hrs 19 hrs 45 58 hrs 10 hrs 26 hrs 53 min 55 min 45 min ! 45 min 15 min 30 min min 30 min 15 min 12 min
- noticed severe loss of flow at FY101, inspected can at 1100, plashc melted down
" Transferrred to drum of water and hardened residue chopped up i t
1
- " Run aborted due to flow at FY 101 down to 0.5 scfm. process can inlet piping filled with water, could not boil off since heat tracing out of commissson, melted sample, blocidng flow.
J
@ Run 7 restarted for extended purge only NA notanalyzed j
~11U RPT--2 03 -
Rev'0 Page 9 Of 87
_______.s. _m m_- . . _ . . - . _ _ _ _ . _ . . _ , _ . __m. _ _ _ . _ -__._._._._______,____.______m_..____m.__.,______ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ m _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ , . _ . _ _ _ _ _ _ _ _ _ _
l 4 1 SEG/TRJ/R-006, Trajan S:c:nd Integrcted Test Data Repart j@@@ '
== mo ~ .c 1.3 Summary of Hydrogen Analysis Results ]
1 The residue H2 analysis performed at the WNP-1 site was methodically conducted because of the critical need to maintain very low H2 and H 2O background levels in the glove box, gas chromatograph (GC), sample bags, etc. A summary of hydrogen 4
analysis results is listed in Table 1.3 on the next page.
l 4
Three " Phenanthrene Blank Spike" samples which involved the addition of a known l hydrogen contem, approximately equal to the NRC critical quantity, in the process )
can, were run to determine recoverability of the quartz tube furnace system. The first and second phenanthrene analyses recovered 78% and 64% of the known hydrogen which are within the expected range per the sampling and analysis plan. The third '
phenanthrene analyses was invalidated when the carbon bed in the quartz furnace was bumped, causing a large upward shift in hydrogen background. A blank run with an empty quartz boat produced a low b4ckground, equivalent to approximately 0.007 i gram-moles (g-moles) of Hi in a process can. However, to be conservative and because it was of !!ttle consequence, the blank was not used to correct hydrogen i content values.
Of the non-PVC test runs (Runs #1, #2, #5, and #8) only Runs #5 and #8 (without the
" cloth") met the hydrogen criteria. Run #1 random sample met the hydrogen criteria; however, the biased sample did not because some of the waste was sheltered from processing when it fell into the center of the Raschig rings. Before Run #2 started, the surrogate waste melted into a hardened mass while waiting in " Standby Mode" at 500*F for more than 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. The hardened mass inhibited diffusion and slowed steam flow rates which greatly slowed the chemical reaction rates. Contributing to the high H2content of Runs #1 and #2 was the grapple leaking water into the residue.
Run #5 produced satisfactory results without Raschig rings, but took too long. The Run #8 residue was contaminated with a small amount of cloth shaped char which caused the initial hydrogen analysis to indicate high. Cloth was used to clean the process can prior to use. The polypropylene filters melted and lost their clothlike appearance with heat applied, therefore, the cloth shaped char could only come from a I cotton based cloth used during cleaning. A subsequent analysis showed that Run #8 did meet the hydrogen criteria if the cloth was not included in the residue sample. l Run #8, with Beil saddles in place of the Raschig rings, proved that a run could be completed in a 12 hour1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> shift and meet the hydrogen criteria. The process requires some form of stainless steel packing shape to enhance the heat transfer rates, and the crude Berl saddle shapes do not have inner ring areas to shelter portions of the waste from processing.
Runs #3, #4, # 6, #7, and #9 contained the same amount of polypropylene sock filters as the other runs plus " black filter flange material"(see Section 4.3 for a description),
which was determined to be PVC, crude oil, and several other additives instead of neoprene. Run #3 was conducted with Raschig rings and 534 g of PVC cut into 1" by 1.5" strips. Run #4 was aborted when steam flow through the process can fell to zero due to excess condensate caused by pipe heater failures, and the waste melted down TRJ-RPT-2-03 Rev.0
SEG/TRJ/R-006, Trdan Sec::nd Integrated Test D:ts Report @@['@
=,- - .m .c into a hardened mass similar to Run #2 as CFE temperatures rose without any steam flow to the process can. Pipe heaters were replaced, and a new condensate drain installed below the process can to drain out any condensate that may be blocking flow. For all runs subsequent to Run #4, procedure changes were made to ensure steam temperature and flow are high enough to prevent a hardened mass from forming before the CFE temperatures are raised. Run #6 was a repeat of Run #3 but without Raschig rings. Run #7 repeated Run #6 bM with a much more stringent end-of-run criterion. In spite of these variations, Runs #3, #6, and #7 did not meet the hydrogen criteria. In fact, these runs produced 30 to 50 times the hydrogen limit without regard to the process variations. After investigating the nature of the "PVC" (see Section 3.26 for details), Run #9 was conducted with the stainless steel Berl saddles,53.5 g of PVC cut into 1/4" by 1/4" pieces, and operating temperatures 100*
hotter than previous runs. Run #9 met the hydrogen criteria after applying a dross c,orrection defined in Section 3.24.
The hydrogen determination for each of the runs has been tabulated in Table 1.3. The maximum allowable hydrogen content in a process can is defined as 0.05 g-moles H i when it is 50% filled and five process cans are placed in a process can capsule. The limit for each loaded process can capsule is 0.250 g-moles Hi .
1.4 End-of-Run Criteria The Integrated Test equipment utilized in-line gas sensors for H2 , CO,0.2a .d benzene displaying near real-time concentrations in the steam-reformer vet.t Wdous gas analysis criteria for defining end-of run were used throughout the tests. For Runs #1 through #3, the criteria was for H2to reach a steady level for at least two hours.
These runs were not successful in meeting the Hi criteria. Runs #5 and #6 used the criteria that H 2must remain less than 500 ppm for the last four consecutive hours of the run. Run #5 was successful, but Run #6 was not. For Run #7, the criteria was that H2must remain at background level for the last four consecutive hours of the run.
Run #7 was not successful due to the large amount of PVC being processed. Runs #8 and #9 used the criteria that H2 must remain less than 200 ppm for the last three consecutive hours of the run. Both runs were successful. After nine runs with varying degrees of success, the end-of-mn criteria which correlated best with a satisfactory residual hydrogen analysis was:
(1) Hydrogen levels measured in the SSR vent line at or below 200 parts per ndtlion (ppm) for three continuous hours, declare End-of Run.
(2) If hydrogen rises or spikes above 200 ppm, the three hour period must be restarted as soon as hydrogen is again at or below 200 ppm.
(3) If available, CO and benzene levels temain at or below background levels throughout (1) and (2) above. (Background level for benzene is defined as less than 0.6 ppm)
TRJ RPT.2-03 Rev.0
SEG/TRJ/R-006, Trojnn Szcznd Integmted Test D:ta R:p::rt [ }@@@m
-cm.m (4) After End-of-Run has been declared, commence the nitrogen purge and cooldown modes to end the run.
h .
Run # Sample PVC in Surrogate Waste Residual Hydrogen Result i Bias No 0.188 g-moles H I Random No 0.036 g-moles H 2A Bias No 0.057 g-moles H 2A Random No 0.192 g-moles H 2A 2nd Bias No 0.134 g-moles H 3 Bias Yes 0.173 g-moles H 3 Random Yes 1.415 g-moles H 3 " White coated lumps" Ye: 1.500 g-moles H 5 Bias No 0,041 g-moles H 5 Random No 0.052 g-moles H*
6 . Bias Yes 2.687 g-moles H 7 Bias Yes 1.650 g-moles H 7 "No White coated lumps" Yes 0.066 g-moles H l 8 Bias, cloth No 0.081 g-moles H 8 Bias Duplicate No 0.029 g-moles H 9 Biased Yes 0.030 g-moles H" 9 Random Yes 0.029 g-moles H**
Sample exposed to glove box atmosphere, value corrected for moisture is 0.047 g-moles H,
- Run #9 values are corrected with the dross correction of Section 3.24. ,
l l
TRJ-RPT 2-03 Rev.0
SEG/TRJ/R-006, Tr jm Sec nd Integrated Test D:ta Report M]
KEWFC ECOLOGY G400P. sc 2.0 INTEGRATED TESTS AT WNP-1 SITE The Integrated Tests were conducted in accordance with the Trojan Second Integrated Test Plan Procedure (SEGfrRJ/ PRO-033). At PGE's request, several runs included "PVC" in addition to the polypropylene filter material specified in the test plan. This section discusses the overall test objectives and results, residual hydrogen and gas sample results for each run, provides comments on achieving the data quality objectives, and describes the results for each individual run.
2.1 Test Objectives and Results There were two objectives for the second Integrated Test. Each of the objectives and comments on the degree of success are listed below:
2.1.1 Confirm that the Trojan production debris processing system can reduce organic filters and dross to less than 0.050 g-moles mono-atomic hydrogen H (<0.05 grams total hydrogen content) per process can. Since five process cans fit into one disposal capsule, this limit equates to less than 0.250 g-moles total mono-atomic hydrogen (0.250 g H i) per process can capsule.
Results: Polypropylene (without any PVC) test runs #5 and #8 (without the " cloth")
successfully met established hydrogen limits using biased sampling and no dross corrections. Run #9 verified that a polypropylene filter batch with 53.5 grams of PVC (which is 10% of one filter top and bottom) cut up into 1/4 inch pieces also met the 0.050 g-moles Hi per process can limit when applying the dross corrections. Therefore, the SSR can reduce organic filters and dross to less than 0.050 g-moles mono-atomic hydrogen (<0.05 grams total hydrogen content) per process can.
2.1.2 Determine the following control limits for operating the debris processing system to assure that each batch processed meets the hydrogen specification:
- a. Determine the length of time, process temperature, and flow rate required to successfully reform a process can full of organic filters and dross.
Results: Runs #1,2,5, and 8 processed polypropylene filter material under varied operating temperatures, flow rates, and processing times. Runs #5 and #8 (without the " cloth") met the hydrogen limit criteria. Runs #3,6,7, and 9 processed varying amounts of PVC added to the polypropylene material at various temperatures and processing times. Only Run #9 met the hydrogen limit criteria. Since production process cans at Trojan may contain some PVC, and because the higher temperatures are considered conservative, the operating conditions established for Run #9 will be used at Trojan:
- Superheater outlet temperature setpoint: ll75*F TRJ.RPT-2-03 Rev.O
SEG/TRJ/R-006, Trc,j:n See::nd Int: grated Test D ta R: port _
j@@@
e Steam inlet to Process Can temperature setpoint: Il50*F e CFE steam chamber temperature setpoint: 1125'F e Steam flow process can: 2 to 5 standard cubic feet per minute (sefm) e Process time to be determined by hydrogen and CO readings as stated in Section 2.1.2.c below.
- b. Determine the minimum diying and cooldown time for the post-processing nitrogen (N 2) Purge.
Results: The hot (1000*F) N2 Purge time used at the end of each run varied from 10 min to 60 min, followed by a 2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> cooldown with 300*F N 2. There was no quantifiable difference in drying results. The shortest drying period used during a succesful run was 25 minutes during Runs #5 and #9. Therefore, the N 2 purge time used at Trojan will be 25 min.
All runs cooled down the CFE using 300*F N2 to 500*F in 2 hrs 20 min.
- c. Correlate in-line hydrogen and CO instrutnentation with measured hydrogen levels to provide process control information.
Results: The in-line gas analyzers displayed sharp peaks for hydrogen, CO and benzene. All peaked nearly simultaneously. The CO analyzer lacked sensitivity at levels below 200 ppm and therefore, always returned to background level well before the hydrogen or benzene levels during the end portions of each run.
Benzene levels also tended to level out at very low levels well before processing was complete. The final "end-of-run" criteria used successfully in Runs #8 and #9 and planr:ed for all Trojan production runs are:
(1) Hydrogen levels measured in the SSR vent line at or below 200 parts per million (ppm) for three continuous hours, declare End-of-Run.
(2) If hydrogen rises or spikes above 200 ppm, the three hour period must be restaned as soon as hydrogen is again at or below 200 ppm.
(3) If available, CO and benzene levels remain at or below background levels throughout (1) and (2) above.
- (Background level for benzene is defined as less than 0.6 ppm)
TRJ RIT-2-03 Rev.O
SEG/TRJ/R-006, Traj:n Secznd Int:gr ted Test D:to Rep::rt m )@@@
MK IM M *C (4) After End-of-Run has been declared, commence the nitrogen purge and cooldown modes to end the run. ;
2.2 Functional Run Description i
Prior to conducting Run #1, a Functional Run was completed to " shake down" the steam reformer system, transfer bell, support systems, and the residual hydrogen .
. analysis equipment. This run included polypropylene fliters and dross residue with Raschig rings. Several procedural and material concerns ivere smoothed out during this run. !
2.3 Run #1 Description The sunogate waste included 1" stainless steel Raschig rings distributed in the bottom of the process cans. Run #1 involved the shonest processing time of 6 hrs 55 min I with CFE temperature at TE-K102 averaging 1029'F. During Run #1, band heaters l were used on the process can inlet piping. Four of eight inlet holes to the CFE steam chamber were plugged to increase steamflow through the process can. Table 1.1 summarizes processing conditions for all runs. The process can was loaded into the CFE the previous night and the it was held at 500*F for 17 hrs. The heat up with steam began at 1219 hrs on May 17, and in 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> reached a CFE temperature of i -800*F, whereupon the hydrogen level increased rapidly to over 20,000 ppm. ' Die CFE reached 1029'F by 1600 hrs. End-of-run was defined as H2 < 500 ppm for at least 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br />. Figure 2.1 shows the correlation of hydrogen and CO for the mn as a function of the elapsed time from the start of the run.
The gas composition at the vent was monitored by the in-line sensors to detect any additional hydrogen evolution as the temperature increased. Both hydrogen and CO evolution increased simultaneously from a few ppm to peak at values above detector saturation. They also declined together at the end of the run, as shown in Figure 2.2.
O2remained relatively steady around 3% throughout the mn. The in-line sensor data suggests that both hydrogen and CO were valuable indicators of processing and process completion.
The metal dross in the surrogate waste did not fuse together, allowing the residue to be easily removed from the process can. Biased sampling of the material shielded in the Raschig rings revealed that Run #1 did not meet the hydrogen limit of 0.050 g-moles H i. Random sampling of material outside of the Raschig rings did meet the hydrogen limit.
TRJ-RPT-2-03 Rev.0
H CC Q tt) )
k O 3 S w 5 CONCENTRAT10h ,
12:19:00 q 12:36:40 12:54:10
[to l C 13:11:40 y 13:29:10 $
M 13:46:40 i g 14:04:10 14:21:40
/ e" 3
l 14:39:10 e" 14:56:40 m 15:14:10 ,
Q 15:31:40 Q
_ p -- -
49:00 ,7 p -
8 ,8
{ 15:
, g 16:06:30 l _ 16:2s- 40 "f 3 e
[ 16:42:1s
$ 16:59:40 17:17:10 17:34:40 g
17:52:10 gn l
j 18:09:40 18:27:10 18:44:40 l
19:02:10 l
l is 8 9*
8 i?
4 C
CMO4 h a i
e
- HhNy EgM$g H@ c 'E g33T '
1 1
- Dm PN 2$ 3 $Q g :C2 D s
6 0
1 8./ ..
.gy T
A g8 R '
N T
N E
T, - ::: . 8i C 8 (
E3 N
O a L C -
u u
8 n,.
o f[
l 0 0 0 0 0 0 0 0 0 0 0 0 01 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1: 2 3 4 5 0 :
0 1 2 3 4: 5 0 1
2 3 4 5 0 : :
1 2 3 : 4: 5:
5 9 3 7: 1: 5 0 4: 9 3: 7: 1: 5 9 4: 8 2 6 0: 4: 9 3 7: 1 5 9 4: 4: : 5: : 5: : 0: : 0 : : : : :
1 1 1 2 2 3 3 3 4: 4: : 5 5 0 0: 0: : 1 1 2: : 2 2 5 5 5 5 6 6 6 6 6 6 6 6 6 6 6 6 1 1 1 6 6 7 7 7 71 1 1 1 1 1 7:
1 7 7 1 1 7:
1 1 1 1 1 1 1 1 1 1 1 1 1 1 dE2E4 $
.M3ab F4. O mm% 4-li ,! l I l1'
i SEGfrRJ/R-006, Trojtn S:c::nd Integrated Test Data Report L )@@@
== = == = - .c 2.4 Run #2 Description i Run #2 was designed to be a duplicate of Run #1 in all respects. However, the loaded process can was placed in the CFE in " Process Standby" mode and held at 500*F for ;
over 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. A visual inspection prior to starting the run revealed that the '
polypropylene melted down into the dross and formed a hardened mass which drastically slowed the steam reforming reactions, ,
Processing began at 1152 hrs on May 19. Very little hydrogen was detected until
{
2030 hrs. CO was not detected until 0120 hrs on May 20, at which time hydrogen !
levels reached detector saturation at 4,000 ppm. CO ne,rer exceeded 4,000 ppm !
indicating a very slow reaction rate. Figure 2.3 shows the gas analyzer data for Run !
- 2.
At 0510 hrs the run was terminated and the process can cooled down, removed, and l
examined inside. The material was fused and adherent to tne container walls. The filter top was removed and the contents broken up by a heavy ramming tool while !
immersed in water to simulate spent fuel pool conditions. Once broken into small pieces, the contents were mixed thoroughly and the process can resealed. The can i was drained and put back into the CFE for reprocessing as kun 2A. I After the process can was put back into the CFE for reprocessing, it continued ;
processing for another 16 hrs and 45 min. CO and hydrogen remained below 400 ppm throughout Run #2A. Figures 2.4 and 2.5 show the gas analyzer data for Run #2A. At 0700 hrs on May 21, Run #2A was terminated and the process can removed for hydrogen determination. Because the residue was fuzed into hard lumps, representative sampling was difficult. Two biased samples varied greatly, and the random sample had the largest hydrogen content. None of the samples met the hydrogen limit.
i l
TRJ-RPT-2 03 Rev.0 p
l
-4 Ch E PPM ff5 k
[ h o b 6 20:14:30 .
6 20:37:20 ,
i 21:00:10 i J
h ch i
/
- l 21:23.00 (f
1 21:45:50 d
9 8.
22:0s:40 22 31:30 g --
g w
22:54:20 CL 23:17:10 4 h=
- E 23:40:00
- 00:02:50 #
h 4 O 3 l 00.25:40 k b s I 00:48:30
,- (m v d
c:
-4 t . . . . . . . . . . .
I 01:34:20
....s
-I D m *fe ~D 88 g 'I 01:57:10 ~ ' ~ *
- p t ._ _
~
H E 02:20:00
., a. _,...
rr #* !
- *'. z e o
02:42:50
'-s*--= t =
m o
D g 2
q 03:05:40 j e. C 03:28:30 Ef.r. . M d^
03.51:20 bcyp ,n...-...
04:14:10 j{"
04:37:00 5 e
('[*t -
04:59:50 ng.y ........ ... a-05:45:30 06 08:20 06:31:10 06:54:00 07 17:10 60h 11- 8 x
R o
CM0 4*C.gihp ]e.hQ yC C . -
]k edU yyk t
$lE e u
- -@gyE O>M g h
h-
- , l N
O I
T A _ G$8 R
T - m 81
[4 N _ k
, ll.
E C
N 51
{
O .
C .
8-
. 9 -
. ! [
- ~
~ % _
r .
o I Y. s 8
0 0 10 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 10 0 0: : :
2 1: 3 ; 4: 5: 0 : : 1 2 3 4: = 0: : : :
2 3 4: 5 0 2 3 3 4
- 5 5 _
5 0 5 6 2 7 2 7: 2 m3: 5 : 72748: 8: 9 : = :9 0: 0 1 1 3 8 3 8 3 8 4 9 4 9 4 9 4 9 2 5 1
4: :
1 3 0 5 1 4: 0 3 5 2 4 :
1 3 0 2 5 1 2'
4 4 5 5 e 6 6 : : : 2 3 3 0 0 1 1 1 1 , 1 1 1 1 1 1 1 ' 1 1 1 1 2 2 2 2 2 2 2 2 2 0 0 : : : : m0 1 m5w$
Hh4$%w o o
ms% o OmO ll i . 1 l(lll lI ; i{, < II
]l \iill o
tmOg%%g .
E a d @ " C L g lC4 m* d M 8 n2 Y L 6 g 09#
4 >
3 l
8 T 'n
- m5Cmm nN s d$)C2 N> $0 om mCZ -
R5 3o
^8 8o O M5 f
I _
T A
R _ @f 8 T
N No : ' : 8 Ea E '
C m 5 Ea N u -
O C 8o "
s u
m -
-8 f
-8
~
8
~ ]
~- :
M o
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0: : 2 4 0 2 4 0 2 4 0 2 4 0 2 4 0 2 4 0 2 : : : 4: 0 2
- : 4: 0: : 2 4: 0:
0 3 6 0 3 6 0 3 6 0 3 6 0 3 6 0 3 6 0 3 6 0 3 6 0 3 6 0 0: : : 1 2 4: 5 0 2: : 3: 4: 0 1: 2: 4: 5 0 2 3 4 0 1 2 4: 5 0 2 3
- : : : 4: 0 1: 2 2 2 2 1 1 1 1 3 3 3 3 3 4 4 4: 4: 5 5 5 5 5 6 6 6 6 7:
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 dE3E45 4Eg
- "6" o
- hO o
m 9 Omo" l! 1 . 1 lll1lli ! l l ll il{ll
SEG/TRJ/R-006, Trden Secand Integrated Test Data Rep:rt uN@@
50ENTric 8COLOGY 44tXA E 2.5 Run #3 Description As a deviation from the original test plan (SEG/TRJ/ PRO-033), PGE requested that PVC be added to the surrogate waste to determine the ability to process this material.
Run #3 was designed to be exactly like Run #1 with the exception of adding one cartridge filter top and bottom (533.7 g of PVC) in 0.8" x 1" x 1.5" chunks. The surrogate waste included I" stainless steel Raschig rings as well as polypropylene filters. The run commenced at 1350 hrs on May 21 and the CFE reached 1046*F.
The hydrogen level measured at the vent by the in-line sensors very quickly increased from baseline to over 4,000 ppm, at which point the hydrogen sensor was valved out to guard against over ranging. The CO detector was changed to a 5 to I dilution ratio so that all values shown on Figure 2.6 between 1620 hrs and 1840 hrs must be multiplied by 5. CO peaked at about 77,000 ppm. At 1840 hrs, the CO detector was returned to normal flow and the hydrogen detector was valved in at 5 to 1 dilution. At 1930 hrs, the hydrogen detector was retumed to normal flow. Hydrogen slowly and steadily dropped from over 4000 ppm to around 500 ppm in 5 hrs. The CO and hydrogen increased simultaneously and decreased together . However, CO decreased to background level well before H 2-The end-of-run was determined to have occurred long before 0840 hrs on May 22 because hydrogen fluctuated between 500 and 1000 ppm for over 8 hrs. Figure 2.7 shows the end-of-run period of Run #3.
At the end of the run, the CFE was purged with hot N 2for 10 minutes and cooled down to 500*F with N .2 This was done with all runs. However, the process can was left in " Process Standby" mode in a N2 Purge for a total of 23.5 hrs. The biased l sample of Run #3 did not contain any of the " white coated lumps". The random l sample and the sample of the white coated lumps produced residual hydrogen levels ten times that of the non-white lumps " biased" sample. These results indicated that the white coated lumps are unprocessed organics. None of the samples met the hydrogen limit.
Because of the moderate temperatures used in this run, the metal dross in the surrogate waste did not fuse together significantly. The residue in the interior passages of the Raschig rings was not fused and easily fell out of the rings with a little agitation. The black filter material caused the reaction rates to be much slower than in Run #1.
TRJ-RPT-2-03 Rev.0
T aE S R
J E .
R 9
2 0
2 0
4 0
6 0
8 0
1 0
0 1
2 0
1 4
0 1
6 0
1 8
0 G
/
0 0 0 0 0 0 0 0 0 0 T- 0 g 0 0 0 0 0 0 0 0 ;
) T .
2-0 d88 _
R J
3 dN8 p /
R-d49 _
0 0
558 ,6 5N8 T r
o
%59 j a
n
$N8 2
$48 n
, d 599 l I s. d}; f,$$>2 t,i*I .
I
_ i.i ly n d%8 ;-
j;{ F t
e htr I
G g WMR ih Y. U r
! : : R a 5dS - .
,A E t
e
.i 2 d 6W8 .
T I
\i y 6 T
e M
E 988 . t :j i,
T s
( E t 4i% . L S k\.
M 9N8 t D
a ;
i'f T y T. a
(
P 2 9ER 4 R t a )
1
,.h U a g 9NS .- N e .
- R e
2 3 5#8 g 3 p
p9 G o o
f N88 A r S t 8
7 NM9 h*, A .
N A
NN8 4 L Y
9 S NNR -3
-. *A . I S
N#S ~_.
r d88 i
i p>
Va_
d58 ds9 N58 N48 --
-~
H c0 2 o2 p p %
p p x .
mm10 0
R e
.v 0 _
SEGfrRJ/R-006, Troj= 2nd I tegrated Test D:tn R: port FIGURE 2.7 TEST RUN #3 END OF RUN e000 a
m ,d R
a
\:1
- g i
4:
9 3000- - --
02% x 100 3 . . co ppm i
i H2 ppm 2000 --
g d.
3,,
f.
1000
'Yi,
't -
'?ep.
w
%O,o L
'7:yng.lI
,r.s + ss ~ " ~ n
-1000 8838898898RS88%8RS8RS8RS8RW88 NdW9sNMdM8N%9NNN#8NNN9$NMdM8N NN99NNN55NddNNN88555dd8ess8ss TIM E(May 21 - 22)
TRJ-RPT-2-03 Rev.0 g gpg7 ,
SEG/TRJ/R-006, Tr jtn Secand Integrated Test Data Report ,_ M@@
a,- .- .e j 2.6 Run #4 Description l Run #4 was temiinated when the failure of the piping band heaters allowed a build-up t
of condensate in the steam inlet to the process can. Initially water from the process can drained into the steam inlet piping. Incoming steam condensed in the cooler I water and added to the problem. Process flow was blocked while the CFE band I heaters continued to heat up the CFE and the process can. As a result, the surrogate waste melted down and reformed into a hardened mass that blocked steam flow completely. A positioning bracket was machined, and all piping band heaters were replaced with radiant heaters. An additional drain was added to the CFE piping to better deal with any condensate collecting at the process can inlet. In addition a procedural change was made to keep the CFE temperature about 100*F less than steam temperature at TE-K101 to ensure that the CFE will not heat up while ;
condensate is blocking the steam inlet or keeping inlet steam too cool for processing. I A test on an empty can proved proper heater operation, and the systen was prepared for the next run.
l After Run #4, it was discovered that the transfer bell grapple tool may have leaked water into the residue of Runs #1, #2, and #3 contributing to the high residual '
hydrogen levels analyzed from these runs. The grapple was modified and a N2heater added to the transfer bell system.
2.7 Run #5 Description Run #5 was designed to test the new modifications and procedures. The run included surrogate waste without the PVC. This run did not include 1" stainless steel Raschig i rings to avoid providing shielded areas which are difficult to process. Run #5 utilized j a 25 minute hot N2 Purge drying cycle after processing was completed. All other l conditions were the same as for Run #1.
Steam was applied at 0456 hrs on May 31 to start Run #5. Hydrogen and CO levels
- began to rise between 0615 hrs and 0645 hrs when processing temperatures were l reached. The hydrogen detector was valved offline at its saturation level of 4000 l ppm at 0730 hrs. It was returned on line at a 5 to I dilution ratio at 1000 hrs. Normal l hydrogen sample flow was restored at 1030 hrs. The CO levels paralleled the hydrogen reading. At 0745 hrs the CO detector was placed on a 5 to 1 dilution ratio I to keep readings on scale. At 0800 hrs the CO detector was valved offline when
- readings exceeded 100,000 ppm (5 X 20,000). At 0830 hrs the CO detector was l returned on line at the 5 to I dilution ratio with readings over 82,000 ppm. At 0940 i I
hrs the CO detector was returned to normal sample flow. Figure 2.8 illustrates the early part of run #5.
CO levels went to background at 1200 hrs on May 31, but hydrogen lingered between 200 and 1000 ppm well into the evening. The hydrogen levels went up and down in cycles which peaked in synchronization with the CFE band heater cycles. As TRJ-RPT-2-03 Rev.O Page 25 of 87 -
SEG/TRJ/R-006, Trejtn Sec2nd Integrated Test D:.ta R: port d@@@
==c == = .c temperatures went up in the CFE, hydrogen production increased. The gas analyzer was off line for cleaning between 0045 hrs and 0115 hrs on Jun 1.
For this run, the end-of-run was defined as 4 hrs at or below 500 ppm hydrogen. This was achieved at 0825 hrs on June I when the run was completed and purging and -
cooldown commenced. Processing time was 27.5 hrs. End-of-Run is illustrated in Figure 2.9. The lengthy time to reach end-of-run is attributed to the formation of ,
large hardened lumps of dross at the bottom of the can. These lumps inhibited diffusion and therefore, slowed the process. Raschig rings prevented this problem in the earlier runs; however, the rings provided hiding places for waste to remain unprocessed and contributed to Runs #1, #2 and #3 not meeting the hydrogen criteria.
Run #5 met the hydrogen criteria. ,
l t
l l
I 4
i i
f l I
TRWT 2-03 Rev.0 P 87
( , _ _
l Ch g PPM g e
o y
04:56.00 (; l 05:19:50 , i 05:43.40 g 06:07:30 '
06:31:20 ,
06:55:10 \ *
~
07:19:00 , j-
~
g 07:42:50 m -;ZEm - ---- =
m:06:40 ..
'N -
Q.
G......._.
g g;g;g <_.-
. . . . . . . . . . . . . . . . . - . .p
""s ~
go 08:54:30 ,,,,_. @ g 09:18.40 / '~~ $ $ '
09:42:40
' f;*7 p C.
J;r. o, q I 10:07:39 .r' * . $
2
- E 10:31:40 C g
,m 5 10:55:30 2
- C to
(
~
% 11:19:40 w
q e un a 11:43:50 g m
L
]
oo 12:08:00 $ 3
" 12.32:00 I.
? $
3>
12 56:00 # r-13:20:00
!! if m
13:44:10 14:00:00 14:31:50 14:55:40 15:19:40 15:43:50 i
5 iS 8 g M
.a 8
m 4
o
l 1
1 PPM g s x e y 8 $ $ 8 Q .
g 22mm 1 . ,
, go
- 22:n 40 . . , . . . . _ . . . . . . . . . . . _. . _m 3 22:47:10 23.10:40 y p j l
23.34:00, , m.{
l M 23:57:20 , ,
l 1 00:20:50. I N
00:44:10 g 01:07:30 01:30:50 7 I d
e 01:54:10i
'f
% m 02:17:30 N' '
b C
4 f n N 02:40:50; m
%y $ H p
g 03:04:10 % e g i e . . i,
- i 03:27:40 ,_ ,
g e
[ g : g b d 04:37:40 '
.( $
o 05:01:00 jd
$ 05:24:20' O
O 05:47:50 \ -
m t
08:11:00 N h 2
l 08:34:20 08:57:40 7 07:21:00 07:44:20 .
w
.= .J==
08:07:50 ,' .
E 08:31:10' '#3 08:54:30
}
l :
i :
1 .
l 58g i
M l 8 l
?
O
SEG/TRJ/R-006, Trojan Second Integrated Test Data Report y [@@
sw- aan -
2.8 Run #6 Description Run #6 was designed to be exactly like Run #5 with the exception of adding one cartridge filter top and bottom (531.6 g of PVC) in 0.8" x 1" x 1.5" chunks. No Raschig rings were added. In addition, a benzene monitor was added to the gas J
analyzer rack. The steam flow through the process can was increased to 5 scfm, which was up from 2 to 3 scfm in the previous run. The process time was 19.75 hrs.
Figure 2.10 illustrates the first 12 hrs of Run #6. Hydrogen levels reached detector saturation at 1500 hrs and returned on scale at 1745 hrs. The CO detector was shifted in dilution factors to remain on scale between 1500 hrs and 1730 hrs. Benzene peaked simultaneously with the hydrogen and CO.
l The end-of-run criterion was the same as in Run #5, hydrogen at or below 500 ppm for at least 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />. During the last 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> of the run, hydrogen showed the same cyclic peaks noted in Run #3. Figure 2.11 illustrates Run #6 End-of-Run.
When the residue of Run #6 was removed from the process can, a few ribbons of l uncooked PVC, similar in appearance to "beefjerky", were noted. These contributed l heavily to the mn not meeting the hydrogen criteria. The dross residue was free-flowing powder with sand sized dross and carbon lumps. l 1
After the mn, visual inspection and testing showed that the new orifice plugs placed in the bottom of the CFE steam chamber to limit flow had caused the process can to have a poor seal with the bottom of the CFE. The high steam flow logged in Table 1.1 as going through the process can, was actually bypassing the process can at an unknown rate. I Run #6 had the highest residual hydrogen content measured during any run because of the incomplete processing of the PVC.
l l
l l
i l
TRJ.RPT 2-03 Rev.0
d PPM M C M 10 54:11 i i i i i i i i i i S W
t-e w
11:18.00 g' 11:41:50 12 05:40 ,r, n 12:29:30 ,
12:53:20 O*
as E
13:17:10 tJ 13:41:00 E g
14:04:50 W D
14:28:40 7 'TI E
30 14:52:30 N(. s -
i ,~ . . . . . . .........................
5 c
y
~,,Wc-- ;
.. . . . . n i 15:1e:20 w :n a pL,. .. . _ m o.
g 15:40:10 r-vs ,. ., ,., . ..,,.:,....
, **5_ w p.,...,...%.,,_
B8 g
q g 16:04:00 ] .ggg. ,
j p 1,.27:30 f --$ > . . . . . . . . . . ...... .......
....r . ._ g g, i m a 16:51:40 ' ' , ,
) ... -f g ,
l .
g b , - * '
' . e t. .:
M C W o 17:15:30 a
-. * %,wv... 2 m w # .....
D '
C 17:39:20 a R O O O "so 18:03:10 -
> A l 18:27:00 12:50 50 i r
19:14:40
- g" 19:38:20 UI 20 02:10 .
20 26:00 ,
20:49:50 21:13:40 21:37:20 l
i I I
l]g" is 8 xR 8 i a
- o 0
o
Er i L !?! .
Ii > 1 . :! ! f h3 CiCMzb d9
- t
- p 4 g $ r"gdg*d9g ago n3E w .* %mt cZ g gO 2C .
$ l g .'
l g$2! 8 M g 58
=s P
P !.
8 e
- s -
mM'jll g,i
~
g - .
- k
=i '
y 1a. ,'m E- c< .
= ', -
.s;q-
- i
,M n m
.d' a d' ~ m O -
m i uS .
m""' m" m mF
-N 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ,
- 2 3 4 5 0 1 2 3 4: 5: 0 1 2: : 3 4: 5 0 1 2 3 4 5 0 1 2 7:
5 0 5 0 5 1 6: 1: 8 1 6 2 7: 2 7 2 3 8 3 8 3 8 4 9 4 4: 0 1: 3: 4: 0: : : :1 : : : 3: : 4 0 1 3 4 0 1 3 4 0 1 3 4: 0: : : 1 3 4: 0 1: 1: : : : 1:
1 2 2 2 2 3 3 3 3 0 0 0 0 1 2 2 2 2 3 3 3 3 4 2 2 2 2 2 2 2 2 2 0 0 0 0 O 0 0 0 0 6 0 0 0 0 0 0 0 eE2E3* 3 .
Yek b xho m wo .
T SEG/TRJiR-006, Trojnn Second Integrated Test Dato Report b. @@@
2.9 Run #7 Description Run #7 was designed to be exactly like Run #6 with the exception of redefining end-of-run. No Raschig rings were added, but one cartridge filter top and bottom (606.4 g of PVC) was added in 0.8" x 1" x 1.5" chucks as before. The CFE orifice size was reduced from 0.5625" to 0.375" by four inserts to force more flow through the interior l of the process can. These new inserts also allowed the process can to seal on the tapered seat.
? .
l As shown in Figure 2.12, hydrogen and CO were slow to rise during the early part of l l the mn. This was probably due to reduced steam flow to the process can. Although l l the process can filters were cleaned, this was the third use of this process can. At !
0150 hrs on June 6, hydrogen and CO levels reached detector saturation. Benzene l was delayed in rising to a peak due to the detector being isolated until 0215 hrs on .
June 6. CO returned to back ground level by 0600 hrs on June 6. Benzene dropped to L 'less than 1 ppm by 0930 hrs and below 0.3 ppm by 1900 hrs on Jun 6. Hydrogen i l cycled between 100 and 1000 ppm until 2100 hrs on June 6. The same characteristic !
peaking noted with previous PVC runs were evident in Run #7. The last hydrogen l peak above 500 ppm occurred at 2045 hrs on June 7. )
i The end-of-run criterion for Run #7 was to record background level hydrogen for one l hour. Then, the process was continued for four additional hours. As a result '
hydrogen read background level for the last 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />. Figure 2.13 shows hydrogen cycling during the latter part of the run on June 7. Processing was finally completed at 0725 hrs on June 8. The last 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> of data are not available for plotting due to a buffer overload in the computer program. It was not designed for 58 hour6.712963e-4 days <br />0.0161 hours <br />9.589947e-5 weeks <br />2.2069e-5 months <br /> runs.
The residue removed from the process can contained many lumps coated with a white substance and some more porous black lumps similar to lumps of dross and carbon found in previous runs. Lumps varied in size from %" to 2.5" in diameter. The white coated lumps are believed to be PVC residue with inorganic coatings. No " beef jerky" strips were evident. The dross was powdery and easily removed from the '
process can.
In spite of the much longer processing time, Run #7 did not meet the hydrogen criteria l even with dross corrections. Separate laboratory testing of the PVC was initiated.
' Results of the laboratory tests are reported in Section 3.26.
Page 32 of 87
illl. Iil
)!<!Ij 1l li l jltl1l 1 I jij l) il 1 l il ii:l! )i i,\ ll
- l!l l o
CMO yhh* d98'3; _
,g y 8@3E~8& HOR Ck3N t
s 6cWm MC hd ME $ hn, yhrdU t t
c A
c y, -
3 g
i j
g g
j
. g.
[3
.j
- y8 M
- ~
- . 8]
P P . i~ -[ - s il
.a m-y - l l j a,E m.
hw 4.
> ' . . I4['
r.
a , .
- 7. * ,- )f I '
- ' #w
[ j(I I tl I o ll' i
0 2
0 00 0 1 5 0 40 3 0 03 0 3 2 0 0 0 00 05:0 40 30:: 0 1
2: 2 0 0 1:0 0:0 05 04: 0 3 03 50 4 0 3: 2: 1 0:
6:
5 9 3 6 0 4: 8: 2 6 0 4 7 1: 5 9 3 7 1 4: 8 2 6 0 4: 6 2 0 5 1 4: 0: 3 5 1 4: 0 3 5 1 4: 0 2 5 1 4 0 2 5 1: 4: 0 2 5 1 4:
7:
0 1 1 2 2 2 3 3 0 0 0 1: 1 2 2 2 3 3 4 4 4: 5 5 6 6 6 7 2 2 2 2 2 2 2 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 -
4Em
- k 3 N g[6w 4.o a~
T>$
Ii'Ili l1 ll lfl Il1' f!' P!
i\I' . ,! , 1! ltlil1l.
ll l I\
Cm ME h @ smdr y"g, gE0}r 8
$ H$.O k
- 3og $ (*
]j" j dk -
$ 0 8 M - ' 8N P -
P IsN 1
l? ,3 X .8 .
j-
~
5 - i' IL [2E I i.
I
- E,j .
, g - -
g " .
s o
Mm
- m u
u n
! = " - "
- , L- -
0 0 0 0 0 0 0 0 05 30 0 0 05 30 0 0 05 30 0 0 05 03 0 0 0 5 03 0 0 5 3 5: : 3 1 5:
5 3: : 5 3 1
1 1 1 1 1 1 1
4: 8:
4: 8 3 8 2 7 2 6 2 6 1 6 0 5 0 4 9 1 8 0 5 0 4 9 8 3 1 4 0 3; 5 2 : 4: 0 3 5 2 4 1 3 0 2 5 1: 1: 2 2 2 3 3 1
4 0 3 5 4: 4:
- : 4 0 3 5 2 1
- 7 7: 8 8 0 9 9 0 0 4 5 5 6 8 8 7 0: m7 6 1 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 d m"Eed -
- h. ks c?
nOo$
o
. - =- -. .. . - _ - - - - _ - . . - -- - _ - - _ - -
SEG/TRJ/R-006, Tr:#n S:cond Int: grated Test Drta Report Lh@@
n,- um= = .c 2.10 Run #8 Description Run #8 was designed to speed up the processing of runs which do not contain PVC.
i Stainless steel Raschig rings were cut in half longitudinally and then welded together l
to form crude Berl saddles. In addition, all steam orifices in the bottom of the CFE l
steam chamber were loosely plugged (plugs are 20 thousanths of an inch less in l
diameter than the orifices) to force more steam to the process can. The process can, used twice before, was cleaned with oven cleaner and a high pressure water system. -
Flow through the process can was significantly improved.
As noted in Runs #1, #2 and #3, the use of Raschig rings or, in Run #8, Berl saddles
, greatly speeded the process. Total process time for Run #8 was 10.25 hrs. Figure i 2.14 shows the entire run. l l The end-of-mn criteria used for Run #8 was one continuous hour with hydrogen ;
l s200 ppm. After declaring "End-of-Run", the process was continued for 2 additional
! hours. if hydrogen peaked above 200 ppm, the 2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> period was to be restarted. The 2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> period did not have to be restarted, so the run ended after three consecutive :
hours with hydrogen s200 ppm. Run #8 end-of-run data is shown in Figure 2.15.
Absent from this run was the cyclic peaking seen in PVC runs.
The residue was powdery with very small dross lumps. The initial biased sample hydrogen analysis performed on Run #8 included a piece of charred cloth-like material. This caused the analysis to indicate that Run #8 did not meet hydrogen l l criteria. It was determined that the cloth was a tom piece of the rag used to clean the J process can. A second biased sample without the cloth-like material met the hydrogen criteria without a dross correction.
i l
l i
/
Page 35 of 87 I , -
l d PPM M C M i
y 19.00:00 llC b
w 19.20:30 g 19:41:00
}
20:01:30 ,
20:22.00 d
20:42:30 '
d.
n 21:03:00 3
i N
21:23:30 3 21:44.00 ,
5 22:04:30 l
't SQ 22:25:00
%~~~ 5 c
g 22:45:30 ~1 . . . . . . .
..; .. ...... :n m
. .... m L
-4 23.06:00 .*S--m,
--' p d i .. a m 23:26.30 g
)D
~
~'***,,;.., + - **-
- * * * * * ~ ~ --
k 23:47:00 tll'll
- L . . . . .....-- u
~
- - -~ =4
>ct y 00:07:20 ;g #
sn C )ll
% $ 0027:50 Z m w
o 00:48:20
. g "lll3 O
0 0 m
oo 01:08.50
" 01:29:20 01:49:50 .g 02:10:20 h
0230:40 5 l
l 02:51:10 03:11:40 03:32:10 03:52:40 04:13:10 04:33:40 i .
I T i5 8 og 01E-a a-
= 8 3
- =
?~
o .
Illll ll! l :l
]
@* g"~> g4.to2 CMg4@4gb h H ?cU9U yt 1
t ggmmEC n h 2m P*8' A zeZ g g0 Q gZ
$ f j
~
~
~ *
$ ~ '
~ '
$ " J -
~
d
- # I
$
- 1I.
F
~ M
" j S.
- . 0*$
y
. . ':-8,3 M
P m.. .
- 5 83 P a..
8 4 a5 b3 s-u.
p' Ea>)
j
.. J
., - E i
l
{j'
, m 5 u
, pg 1
~
. ..~
o 5 0 0 0 0 0 0 10 05 0 0 01 0 503 0 0 0 0 0 0 0 g 0 0 0 0 0 0 0 3 5: : 3 0: 4:2 : 0 4:2 : : 6. 3 1 5 3 1 0 4 2 0 4 2 0 2 : : : : : : : :
7 7 6 6 6 5 5 0 4: 4: 3 3 3 2 2 2 1 1 1: 0 0 0 9 9 9 8 8 8 5: 0: 2: :3 4: 5 0 1 2 3 0 0 1: 2 3 4: 5 9 1: 21 3 41: 1: 5 0 '. 2 3: 4: 1 4:
0 0 0 0 0 0 0 1
1: 2 2 2 2 2 2 3 3: 3 3 3 3 4: 4 4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ,
dEmp5 aS c
- hC gwhA6w .
mga w4 o%$
l l i j!ll 1 .l \ , .lI(
. - . - . - - . - - - ~ - -...-.. - . _ - -.--- -.- . - .. . - . - . - _
I l
SEG/TRJ/R-006, Trojan Second Integrated Test Data Report [,d@@@
2.11 Run #9 Description -
Run #9 was designed to test the ability to process lesser quantities and smaller particles of PVC. The surrogate waste for this run contained 10% of one cartridge filter's top and bottom (53.5 g of PVC) cut up into 0.25" x 0.25" x 0.25" pieces. The Berl saddles used in Run #8 were also added.
l Some equipmem modifications were made to increase operating temperatures in the CFE. A steam booster heater was added between the boiler and the superheater. This i allowed the superheater outlet, TE-K109, to rise 100* F to 1167*F average. Piping radiant heater element and CFE band heater element temperature limits were raised .
- 100*F to 1400*F, the manufacturer's recommended maximum. This allowed operating the CFE at an average temperature of 1128*F, more than 80 F higher than in Run #7.
l At the start of processing, heat up was controlled such that CFE temperature, TE- ,
K102, remained '75
- to 100* less than the steam inlet to the process can, TE-K101.
This is the same method as used in Runs #5, #6, #7 and #8. All other operating conditions remained the same as in Run #8. Figure 2.16 shows the gas analyzer data for the first 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> of the 26.2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> run.
l The same end-of-run criteria used in Run #8 were applied to Run #9. The last 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> t i of Run #9 are plotted in Figure 2.17. The cyclic rising and falling of hydrogen levels l- was again present. The run took less than half of the time Run #7 took and processed l more completely. This run met the hydrogen criteria when a correction for the dross
, content was applied.
i l
i 4
1 Page 38 of 87
l o l 9 l o:
8 8 x
"]
=
a
$58? ;s
!I 01:85:22 01:43:22 01 0122 01:64:12 01:22:12 01:85:02 IU y 01:43:02
- r 01:01:02 01:64:91 01:22 91 1 @ k
] $ j'l 01:85:81 $
% j LJ ji 01:43:81 ~
=
sa D: . go
@sa 4" 01 01:81 1
=
>MM C 01:64:71 I 2
01:22:71 E O * .,, '~~' ' " ' ~ "
- q s-P a r-..
y;;d. .,-_ -
w4~ - a- w -. .
01:85:61 y
atr L' m,
D: '^ ' ' ' ' ' ' '- ' ^ ' O " b",- -,,,_, 00 43:61 kog h g-
,' _" " '..". u'. 'b aM OW
. . . . . . . 4. . . . . -.
1~ OI E -k -
00:64 51 25 m 00:22:51
% 00:85:41 3
Jt 00 43 41 U
00:01:41
+f 00:64 31 P
C 00:22:31 00 85:21 D[:
00:43:21
%c /
00:01:21 b
% o g u MPP
o d
u 8
8 5 hkbg 8 9 8 a.
s s
Ot:CL:tt 09:00:PL
' 09:LV:CL 00:9C:CL .
p OLZE:CL I g i 03:60:CL
= . 1 i
00:9924
- u. Or:CV:CL 0 \
g, I 09:0CZt g g l
~
,' 00:9LZt %
[ asi f.
06:9025 $o k b E
E g ) OEZS:tt co O p*
3 OC:6C:ti "
b**
- I Or:9z:56 E s 09:C6:46 h b ~
00:LO:L L OL:09:0L E $2 u.
J 03:9C:0L tu)
G OCZE:0L m
j- :
i 09 W 60 l _ ,
00:tt:60 0 L:tC'.60 g 03'8 L:60 0C:90:60 C, 0C:Z9:00 g
km ,,
E ,
OP:6C:00 m 9
ggl I i h ot'93:00 '?
a $ $ k 8 8 8 m O - -
A w
M Ndd h
SEG/TRJ/R-006, Trajrn Secznd Int:grcted Test D:to R:psrt th@@
SODmPC #COLOGY MOuP. eC.
2.12 Data Quality Objectives Data objectives listed in the Test Plan (SEG/TRJ/ PRO-033) are provided with comments on the success achieved during testing. Temperatures, pressures, and flows are reported for Runs #8 and #9 since these were the most representative of planned production operations. Run #8 typifies a non-PVC run. This was the most successful run which utilized the target operating parameters listed in the Test Plan. Run #9 typifies the PVC run. The elevated temperatures and the addition of PVC were the variances in Run #9 from the Test Plan.
2.12.1 Weight of Prototype Process Can Upper Filter Assembly t>efore and after processing to determine how much,if any, of the residue has been captured in the filter: The antic.ipated weight of the upper filter assembly is 4 to 5 kilograms (kg).
Results: During all nine runs, the before processing weight of the prototype process can upper filter assembly varied between 4237.2 and 4301.9
- g. The mass of residue captured in the filter during processing varied between <1 g and 134 g.
2.12.2 Weight of surrogate waste filters: Sufficient filters will be selected and weighed to obtain a weight of 0.96 kg 157 g. (Note that the amount of filter materal and dross weight was selected to emulate the conservatively estimated 1 to 4 ratio of organics to dross expected from the spent fuel pool debris).
Results: For each run, seven filters were selected as surrogate waste. The seven filters filled the process can to capacity. The total mass of the filters used in each run varied between 992.8 and 1016.4 g.
2.12.3 Weight of surrogate waste dross loaded: Sufficient surrogate dross will be weighed to obtain a weight of 3.71 kg 85 g.
Results: Actual surrogate dross loaded in each run varied between 3.6356 and 3.7108 kg.
2.12.4 Weight of residue in sample container after processing: Residue remaining after each test run is anticipated to consist of the dross material I loaded into the process can prior to testing plus a few ounces of processed filter residue.
Results: Actual residue weights varied between 3.642 and 4.080 kg.
2.12.5 Temperature of steam entering the lower inlet of the process can, TE-K101: Steam temperature entering the process can will be 1000*F 100*F.
TRJ-RI%2-03 Rev.0
SEG/TRJ/R-006, Trohn Sec:nd Integrr.ted Test Data R: port _ h@@
Results: Steam temperatures at TE-K101 met test objectives during Run #8.
Values varied from 984 to 1078'F, averaging 1034*F. Steam temperatures were elevated during Run #9. Values varied from 998 to 1174*F, averaging 1137"F.
2.12.6 Temperature of CFE steam chamber, TE-K102: Temperature in the CFE steam chamber should rise to 1000*F 100*F.
Results: Steam chamber temperatures at TE-K102 met objectives during Run
- 8. Values varied from 924 to 1041*F, averaging 974*F. Steam temperatures were elevated during Run #9. Values varied from 1024 to 1183*F, averaging 1128*F.
2.12.7 Temperature of Syn-gas exiting process can, TE-K103: Temperature of the syn-gas exiting process can should rise to 900*F 25'F.
Results: The temperature of gases exiting the process can at TE-K103 were about 100*F cooler for Run #8 than the test objectives predicted.
Values ranged from 773 to 810*F, averaging 793*F. These temperatures were lower than expected primarily due to heat losses from pressure drops across the flow orifices and the process can.
The lower temperature at TE-K103 did not affect the outcome of the process. With the elevated temperatures during Run #9, values varied from 901 to 958'F, averaging 950*F.
2.12.8 Pressure of steam entering bottom of process can, PT-101: Pressure of the steam entering the bottom of the process can should be +22 inches water column (in w.c.) 22 in w.c. when steam is applied.
Results: Pressure at PT-101 met test objectives during Run #8. Values ranged from 18.7 to 22.2 in w.c., averaging 20.1 in w.c. During Run
- 9, a higher pressure drop across the cesium trap caused PT-101 to rise higher than expected. The higher pressure drop is due to fines building up on the cesium trap one micron filter. Subsequent flow tests proved that cleaning with oven cleaner and a high pressure water sprayer lowered pressure drops to Run #8 values. In actual processing, the cesium trap will be replaced if excessive pressure drops are encountered. Run #9 values ranged from 30 to 40 in w.c.,
averaging 34 in w.c.
f 2.12.9 Pressure of CFE steam chamber, PT-102: Pressure in the CFE steam chamber should be between 20 and -10 in w.c. when steam is applied.
Results: The CFE steam chamber steam inlet orifices were significantly reduced in size with the loose fitting plugs. This reduction was done to force more flow through the process can. As a result, PT-102 no TRJ-RPT-2-03 Page 42 of 87
1 SEG/TRJ/R-006, Tr:j:n See:nd Int:grcted Test D:ta Rep rt @@@
longer reads steain chamber temperature since it is now upstream of the flow limiting orifices in the CFE. For Run #8, values ranged l from 24.6 to 28.8 in w.c., averaging 25.5 in w.c. Run #9 values f ranged from 37 to 44 in w.c., averaging 41 in w.c.
l 2.12.10 Pressure SSR inlet: Pressure at the SSR inlet, PT-7 should be between -10 and -20 in w.c. Pressure at PT-7 must remain negative at all times during processmg.
1 Results: Pressure exiting the CFE and entering the SSR was more negative than expected. This was primarily due to higher pressure drops across the cesium trap than expected. Run #8 values ranged from -
16.1 to -32.5 in w.c., averaging -22 in w.c. These more negative values required the SSR blower to operate at higher speeds than planned but well within the system capabilities.
1 2.12.11 Pressure at Blowerinlet: Pressure at the SSR blower inlet, PT-8 should be between -10 and -45 in w.c. when steam is applied. Pressure at the SSR blower inlet must remain negative at all times during processing. .
I Results: For Run #8 values ranged from -22.3 to -28.6 in w.c., averaging 25.1 in w.c. Run #9 values ranged from -13 to -37 in w.c., averaging
-29 in w.c.
2.12.12 Flow rate of steam entering bottom of process can, DPT-101: Flow rate of steam entering bottom of process can should be between 1.5 and 5 scfm.
I.ess than 1 scfm indicates a plugged process can upper filter.
Results: Flow rates at DPT-101 dropped below 1.5 scfm during Runs #2, #4, and #7. During Runs #2 and #4, flow was restricted by hardened j dross due to melting of the plastic wastes before steam reforming i began. Procedure changes are in place to ensure the CFE steam chamber remains cooler than the incoming steam until the system is l fully heated to processing temperature. This method was proven successful in Runs #5 though #9. Run #7 had reduced flow due to reusing a process can without effective cleaning of the 5 micron filters. For Run #8, values ranged from 3.1 to 3.6 scfm., averaging i 3.3 scfm. Run #9 values ranged from 3.7 to 4.6 scfm, averaging 4.1 scfm.
l
, 2.12.13 Temperature of superheater outlet, TE-K109: Temperature of superheater
! outlet should be between 1000*F and 1200*F.
Results: Thermocouple TE-K109 failed during Run #5. The replacement thermocouple had not been calibrated. After Run #9 was completed, the replacement thermocouple was as found calibrated at the cenified Rev.0 l TRJ.RPT-2-03 Page 43 of 87 l
SEG/TRJ/R-006, Trogn S;c:nd Integr ted Test D:ta Report )
__M_. ._
laboratory. It did not require adjustments. For Run #8, values ranged from 1011 to 1130*F, averaging 1061*F. For Run #9 a steam booster heater was installed between the boiler and the superheater.
The booster heater raised the temperature of the steam entering the superheater 200*F. Run #9 values ranged from 1091 to 1178, averaging 1167'F.
2.12.14 Steam pressure at the superheater outlet - PT-105: Pressure of superheater outlet steam should be 3 pounds per square inch gage (psig) 2 psi.
Results: Pressure at the superheater outlet was slightly lower than expected because pressure drops across the flow measuring orifices was less than anticipated. For Run #8, values ranged from 0.39 to 1.23 psig, averaging 0.77 psig. Run #9 values ranged from 1.44 to 1.61 psig, averaging 1.61 psig.
2.12.15 Hydrogen at SSR vent: Hydrogen should track from background at the start of processing to a peak value and back to background level at the end of processing. The curve should correlate with the measured CO curve.
Results: Hydrogen correlated very well with CO levels. The hydrogen analyzer provided a reliable means of measuring end-of-run.
2.12.16 CO at SSR vent: CO should track from background at the start of processing to a peak value and back to a relatively constant level near background at the end of processing. The curve should correlate with the measured hydrogen curve.
Results: CO conelated very well with hydrogen levels. Due to the lack of sensitivity of the CO monitor, CO always went to background level well before the hydrogen. CO levels above 200 ppm indicates that the process is not complete, but CO alone cannot be used to determine end-of-run.
2.12.17 Oxygen at SSR vent: O2 at the SSR vent should be 3% by volume 3%.
O2must remain below 6% to prevent oxidation reactions.
Results: 0 2levels were initially at about 3%; however, a leak in the gas analyzer sampling line was repaired and subsequent runs showed O 2 at less than 1%. During many of the runs, O 2was scavenged to
! <0.1% during the height of the steam reforming and returned to on-i scale readings near the end-of-run.
2.12.18 Syn-gas analysis for H2 , CO, and carbon dioxide (CO2 ) simpled at CFE outlet: Syn-gas analysis for H 2, CO, CO 2by gas chromatograph should range from background at the start of processing to a peak value and back l
Rev.O TRJ.RIrr.2 03 P 4 d 87 I
SEG/TRJ/R-006, Troj:n Sec nd Int: grated Test Data Report .. d@@@
======..c f
to a relatively constant level near background at the end of processing in )
parallel curves. H 2could be as high as 10% at peak value with CO and CO2 Peak values below that of H 2-Results: The Tedlar bag sampling method was not consistent enough to get l reliable analytical data. Since the in-line gas analyzer ws providing j consistent data for tracking H 2, CO, benzene, and O2, Tedlar bag sampling was discontinued.
2.12.19 Syn-gas analysis for O2 sampled at CFE outlet: Syn-gas analysis for O 2-sampled at CFE outlet shculd be <3% by volume 3%. !
1 Results: The GC could not be set up to measure O 2 . O2 was monitored with j
the in-line gas analyzer.
2.12.20 Residual mono-atomic hydrogen in process can residue: Residual mono-atomic hydrogen in process can residue must be <0.05 g-mole or <0.05 grams total mono-atomic hydrogen content.
Results: Residual hydrogen content analysis results for two non-PVC mns, Runs #5 and #8 (without cloth), and one PVC run, Run #9, met the hydrogen criteria.
I i
l Page 45 of 87 1
SEG/TRJ/R-006, Tr$n Sectnd Integrated Test Date R: port LM]
$0ENfFE ECOLOGY W eC.
3.0 RESIDUE TESTING AT WNP-1 FACILITY Residue testing at the WNP-1 Facility involved eight test mns plus several blank and duplicate tests for quality assurance.
Runs prior to 6/4/97 were sampled with a gas tight syringe and injected into the GC. In some instances, data scatter was significant. In runs completed after 6/4/97, the GC was )
operated in the continuous automatic sample mode and data scatter was less than 20 ppm.
The quanz tube purge rate of argon (Ar) ranged from 14 to 44 milliliters (ml) per minute (min). The glove box purge rate of Ar was 1100 ml/ min. The baseline curve for H2 (during the period prior to time 0) typically ranged from 600 ppm to 800 ppm, with overnight background reaching as low as 600 ppm. The determination of the H2 evolution background prior to time 0 was important to establish the shape and slope of the baseline l curve that needed to be subtracted from the residue sample hydrogen evolution curve. In most cases during the daytime, the background curve slightly sloped upward, whereas at -
nighttime this background curve usually sloped slightly downward.
The following sections describe each of the runs and their results. A summary of the data and results for each run is provided in Table 3-1.
3.1 Quartz Tube Apparatus A sketch of the quartz tube fumace and the gas flow tubing placed inside the argon inerted glove box is shown in Figure 3.1. At the outlet end of the quartz tube, within the heated zone, a plug of carbon chips was packed between two wads of quartz wool.
This carbon plug converted any water present in the sample into hydrogen gas by means of the steam-carbon reaction.
3.2 Gas Chromatograph Calibration H 2 A five point calibration of the GC was performed in accordance with the Trojan Integrated Test Sampling & Analysis Plan SEGffR1/R-032. Each day a five point 50, 100,200,500,3000 ppm calibration check was performed. All calibrations and calibration checks ran within tolerances of the sampling and analysis plan.
I Rev.0 TRJ-RfT-2-03 g
i a y
a
- f. l 1 il l l '
!! } 1 3' 'I I y y 9
s s
. a. > . . - -
J 11 l l} It I- t- t g
O
_ m n i s 1 i s i n i i s s 1 s i s g g ,i i gg 3ll3 . . . .
iijl--
3 5 5 5 5 5 5
- 8 i !. !
8 i
. .i i I
. . .! .! .I i B
. .i i l ,
_ijj i
~
5 Is 5 !. I. .i !. !. !. !
. .i .i .i .i I
. .I .i I i l 2 2 9 9 9 f f f f f f f f f f f I E 9 9 9 9 9 Jg g l ((j E ! ! ! ! 1 ! I ! ! ! [ [ g 2 E [ g I ! 5 ! !
gjij ! 5
! i 28 i i ! ! ! 5 i 8 3 5 5 !
- B f ! 8 l
! jo ij3 c ! n ! 5s ! n i ! ! : s i a i ;
a s i a s y
g m
j 3 ! ! ! l 5 ! ! ! ! I I l l ! ! l l I E ! 6 I j g I {j E l @ E I s S I & I ! I E 5 I I E @ n E I l E
! ! ! " i ! i ' ' l ! ! ! !
- l E s
- ! ! ,a
.{ 2 l)8 t
u
*y , )l s
3 e a 2 y a a e a e a a a a e e a s a 3 a y a e a
=
a x y f 1 l - l z [ 33 5 5 i ! 5 ! ! ! g g a a g g s e a 3 y a a g 3 o e
. D- 1g. 1-8 E 8
5 E. ! 8 i E ! !. ! !. ! .E : a
- 2 8
- i. s I
l 3 gLa 8 5
! 8
! E 5 ! S f 5
" 8 8
! 1 I I 8 8 I Ils I
g o a a jfK l l I a a a a a s a a a a j = 8 13 e e a e a a e a s a a II
,i }i q a I I
.& I
" S i S i i " 5 " 8 " ' 8 i
5 E i E 5 i i I}
}jlt R E E E i S R 9 E E E S E R 5 3 E ! E i A i u}l{1g fj
[
. o
=
. i e : s ; ; ; ; ; ; i r
s ; e
; ; -i e
; : m i ^-
E c . . ges: 6 r g t j E j j j j j j j jj [ [ [ j j j j g g j'
- -] a u. ,l g g g E E i g g B j j j j j E 5 g g g' g i E g
E j ;g , u l 5
- S S S $ $ 5 S E S S S 5 5 5 I
s{ a I f
- I ! !, ! I l ;l l h.I= j .i } } ) l ! !s ! ]s ls i
3 n c e a l I > lIIL 11 l_ _ _
SEGfrRJ/R-006, Troj:n Second Etegrated Test Data Rcport FIGURE 3.1 GLOVE BOX ARRANGEMENT FOR RESIDUAL HYDROGEN ANALYSIS I CAYSTER BAcaNc nmcE ci .- eOx ., , i ,
' ' - ~ ~ - ' - - " - -. _ . . - . , _ . . _
uETER [~. _ '"'~._"~.~.~_'.._".._~.._l [~~".._".._~~~.._".._".._'*.._~~.."]{"3"..q*.._..- e l
- lll 7 yl -
l REACTOR TUBE 1
- l
- l 1:
l l::l:
- l ll:
I i Ilti ftXTURE
- y l
l* j' lll lj
^
4 -n. l lllIll . l j $URM(E l
- ,h : -~
l BLUE u IUBE. f. I II I
- l l: :
ll m,L, l::ll R .. s
- l-l*
l :l .
- ll:
.: .._.._.._.._.._.._..J: ; l_ rtOw uETER _y ,_. l _ .__,,__..j-
{l_. . . _ . . _ . . _ . _ . . _ . . _ . . _ . . J L . . _ . . _ . . ._ . . _ . . _ L. .. _. I _r_. .. __. . _] : :L .L [.. _ . _ . . _ . . _ _ . 7 ...__.. _ .. _... _ ..
._y.
.7.._.._.._.._.._...
.7
- q. __y._
. .. . g . . .
l *
. :(
8
'9. y* .
'y y*
.. BEAKER HOUSING
'9. :l
' CC AND ii i,i
,i ;I ,
LLIRA HIC 1 tLTRA HIG t TEDLAR GAC gg !! !! *l Pt'RITY ARCON l'URITY AfiGON ** l: :: I* ll ll ll I! l' i.i 11 ll I! ll ll :*l l 1
! ! !! :: :: i i ::
.. .. : .. . . ....: y y y . .
I..l TRJ.RPT.2 03 Rev.0 Page 48 of 87
J SEG/TRJ/R-006, Trojan Second Integrated Test Data Report E@@
===-m,.
1 3.3 Sample Blank Just prior to heating the sample blank (empty quartz boat), a background curve, shown to the left of time zero in Figure 3.2, was obtained for H 2 flowing out of the quartz tube indicating decay rate and data scatter. The procedure involved removing a new, clean quartz boat stored on desiccant. The quartz tube inlet end 71/60 joint was removed, and the empty boat was inserted tojust inside the heated zone of the fumace. With the boat in the tube, thejoint was replaced as quickly as possible, and the tube sealed again. The heating of the empty quartz boat quickly began as well as gas collection in the Tedlar bag. The hydrogen evolution curve is shown in Figure 3.2. It was the same shape as the other sample runs, clearly showing that the glove box water content, measured by MSA monitoring tubes at.500 ppm, adsorbed onto the quartz boat and entered the quartz tube during sample insertion to produce a blank hydrogen curve. This curve is a true blank arising from handling technique only. The integration of this curve produced a hydrogen equivalent of 0.0070 g-moles H i for a typical process can surrogate waste loading of 3700 grams. At 302 min into the run, the curve had returned to near-baseline. Figure 3.2 Blank Run 840 820 h '
, 800 l' At. . +
~
Z I b** 780 760 -- J 740. I 720 I 50.00 0.00 50.00 100.00 150.00 200.00 250.00 300.00 350.00 Run Time (min) The experimental uncertainty in establishing this baseline curve depends on the extent the carbon bed is disturbed as the result of the quartz tube opening, the insertion of ! the boat, the rescaling of the tube, and any change in At throughput. The hypothesis ! is that the carbon bed was disturbed, exposing fresh carbon surfaces which increased the evolution of H2 , CO and H2 0. Using the error in H2 at 20 ppm the uncertainty in determining this curve is estimated at 0.001 g-moles H i. l Another measure of the " blank" was made by "rebaking" Run #8, allowing the processed residue to be exposed to glove box moisture for 6 hours and the " acquired" hydrogen measured. The result was 0.005 g-moles H i - further validating the blank quartz boat run. The "Rebake of Run #8 hydrogen evolution curve is shown in-Figure 3.3. Page 49 of 87
l 1 SEG/TRJ/R-006, Trojan Second Integrated Test Data Report - x== m= ,. Figure 3.3 Run 8 Sample Rebake Analysis 800 760 ,-
.u 0'
n, 720 N % ) i 700 s l
' - l
! 680 i __ --
660 -- ~ _-
-100 0 100 200 300 400 Run Time (min) 1 1
3.4 Trip Blank l For this. test, analytical grade anhydrous Na2SO, was used as a control, carrying this ' sample throughout all of the handling steps in the procedure including processing in the CFE. This is designed to pick up the maximum possible moisture from the handling steps. The Na 2SO, trip blank of 6.60 g was loaded into the quartz boat within the inerted glove box. The trip blank hydrogen evolution curve is shown in ' Figure 3.4. Just before the sample heating began, a background curve around 886 ppm was obtained for H 2. A decay rate and data scatter of *20 ppm was observed. Figure 3.4 Sodium Sulfate Trip Blank - 25000 l 20000 b E 15000 i k" 10000 3000 / A 0,wd % . . . . . . . . l 0.00 VA.00 200.00 300.00 400.00 500.00 600.00 700.00 RunTime (min) Once the baseline curve was determined with the average of 586 ppm over 170 min, the quartz tube was opened and the loaded boat inserted, at which point the Na2SO, , and quartz boat began heating quickly and gas collection in the Tedlar bag began. The run duration was 671 min. At 233 min a peak began, rising to 23000 ppm, extending over 100 min, and then falling to 526 ppm. The quartz tube purge rate of Ar was 28.5 ml/ min. The glove box purge rate of Ar was I liter / min. Gas sampling occurred once every -5 min. The tail end of the curve leveled out 300 ppm less than the initial baseline indicating a background shift. The 886 ppm baseline value was used for the calculations. After 671 min, the boat was removed from the tube. The validity of the data collected from this sample comes into question, based on two factors. The first concern is that the sodium sulfate decomposition produced H 2S, this was determined by the characteristic smell emitted in the reactor exhaust. The second Page 50 of 87
SEGfrRJ/R-006, Trebn Sec:nd Integrated Test Data R: port I,. j @ @ @ SCSNMC ICOLOST egow, ar. concern is the -200 minute delay in the hydrogen peak. In all the other runs, the peak occurred within 20 min of the sample being placed into the furnace. These factors lead us to hypothesize that a chemical reaction had occurred while the Na2 SO4 was processing in the CFE, forming H2 S, thereby invalidating the data collected from the trip blank. Subsequent satisfactory mns indicated that the method handling was precluding sufficient moisture; therefore, the trip blank was not repeated. The trip blank was not factored into calculations for any of the runs. 3.5 Hydrogen Quantitation Via Phenanthrene Sample This section answers the questions whether the 1100*C quartz furnace, with Ar sweep gas, can quantitatively simulate the total hydrogen production expected from the long-term radiolysis in the repository. Phenanthrene (CuHi o), a heavy polycyclic hydrocarbon (boiling point 310*F), was selected to simulate the residue remaining after steam-reforming because it is available as a certified analytical reagent, the hydrogen content is known, and the material is similar to partially steam-reformed substances. The quantity placed into the quartz boat was selected so that the H 2 evolved would be close to the critical NRC level 0.05 g-moles Hi . To determine the critical concentration of H2 in the sample bag, the ratio of 15 g surrogate waste sample in the quartz tube to 3652 g of surrogate waste residue in the process can and 5 liters of Ar gas per sample bag were used in the following calculations: 0.05 g-moles H x 15 g x 24.5 L Ar x 1x10* ppm 3 = 506 ppm H2 in bag 2H i/H2 x 3632 g x 5 L Ar g-mole Ar Three phenanthrene runs were completed. The recovery of H2 generated from the first nm of phenanthrene was measured 0.000104 g-moles H i for a 4.0 mg sample initially containing 0.00011223 g-moles Hi . From the Sample Blank analysis, background was 0.0000161 g-moles Hi . The first run had a recovery of 78%. A small amoun.t of sample may have been spilled on the scale causing the recovery rate to read slightly low. The second run of phenanthrene measured 0.0000611 g-moles Hi out of 0.0000701 g-moles H ifrom a 2.5 mg sample. The second run had a recovery of 64%, using the integrated value of GC analysis over time since only one Tedlar bag sample was taken. The third run of phenanthrene measured 0.0002702 g-moles Hi out of 0.00020759 g-moles Hi from a 3.7 mg sample. The third run had a recovery of 244%. The reactor tube was jostled when the sample boat was placed into the tube, causing a large upshift in background, invalidating the run. The first phenanthrene run is plotted in Figure 3.5. Runs #1 and #2 met the data quality objective recovery range of 60% to 140%. The curve for Hz evolution is very similar to the residue sample discussed in the following sections; thus, it is concluded that the phenanthrene was a good surrogate for validating and quantifying the H 2 recovery. TRJ-RPT-2-03. Page 51 of 87
1 I SEG/TRJ/R-006, Trojan Second Integrated Test Data Report ggg smus .c d Figure 3.5 Phenanthrene Run 1 Analysis i li 1
. . 1 h* I 1200 7km g'L L,, o W .
.my 900 0.00 100.00 200.00 300.00 400.00 500.00 600.00 j
. ri . war - )
i
. 1 1
3.6 Run #1 Residue Biased Sample l The initial quartz tube H2evolution averaged 523 ppm following an all-night baking : at an Ar purge rate of 26 ml/ min. From -34 min to -4 min (in Figure 3.6), the H2 level I dropped about 20 ppm, taken as a stable baseline. Approximately 10.90 grams of residue was taken from Run #1. It was sampled by j biasing toward the centers of large chunks and materials that appeared visually to be ; poorly processed then placed into the clean quartz boat. At -4 min, the tube was . I opened, the boat was inserted and the tube closed 1.5 min later. Figure 3.6 shows the H2 data as a function of time. After 423 min, the H2level was at an ending baseline value of 1145 ppm and the run was stopped.10.35 liters of gas was collected in the bag plus sampling volume; however, the calculated volume of 12.06 liters was used in calculating hydrogen content of 0.1883 g-mole Hi . This run was above the acceptance value of 0.05 g-mole Hi . The sampling handling blank correction of 0.007 g-moles was not subtracted from this result. Figure 3.6 Run #1 Blased Sample Analysis 1500 m 1000 *
' z ' '
y ' .y ;V V t l l h ; l l 0
! -100.00 0.00 100.00 200.00 300.00 400.00 500 00 !
l Nn Tim e (min) ! l TRJ.RPT-2-03 Page 52 of 87
SEG/TKJ/R-006, Trojan Second Integrated Test Data Report $M@ 3.7 Run #1 Residue Random Sample The initial quartz tube H2baseline was 731 ppm following an all-night baking at an Ar purge rate of 28.5 ml/ min. From -380 min to -4 min (in Figure 3.7), the H2 level dropped about 20 ppm, taken as a stable baseline. Approximately 13.00 grams of residue was taken from Run #1. It was lightly ground by spatula, thoroughly mixed, and uniformly sampled to composite a representative sample. Then this sample was placed into the clean quartz boat. At -4 min, the tube was opened, the boat was inserted and the tube closed 1.5 min later. , Figure 3.7 shows the H2 data as a function of time. After 383 min, the H2 level was at an ending baseline value of 865 ppm and the run was stopped.10.87 liters of gas was I collected in the bag plus sample volume; however, the calculated volume of 11.17 liters was used in calculating hydrogen content at 0.0362 g-mole Hi . This run met the acceptance value of 0.05 g-mole Hi . The sampling handling blank correction of i 0.007 g-moles Hi was not subtracted from this result. l i Ii Figure 3.7 Run #1 Random Sample Analysis I 1400 ! ll 1200 ' A' ' '
! ! I i 1000 j, _ I !
11 :nc t ; ; - ,' ... . __ ----
, I p ( 300 :
A t t l 1 1 I I J i 400 200 i '
' [ !' !' !
l i l
. -150 -100 50 0 50 100 150 200 250 300 350 400 :
Run Time (min) [', Ii 3.8 Run #2A Residue Blased Sample The quartz tube H2 evolution was 878 ppm following an all-night baking at an Ar purge rate of 14 ml/ min. From -84 min to -4 min (Figure 3.8), the H2 level only dropped from 887 to 863, or 20 ppm / hour. This baseline (average value of 878 ppm) was easily extrapolated for the duration of the mn. These data are shown in Figure 3.8. Approximately 13.90 grams of residue was taken from the reprocessed Run #2A. It was biasly sampled for organic char residue in the centers of any chunks. This sample was then placed into a clean quartz boat. The tube was then opened, the boat was inserted into the hot zone, and the tube closed. Figure 3.8 shows the H2 data as a function of time from the start of mn. After 664 min, k H level2 was close to the extrapolated baseline and the run stopped. 9.37 'DU.RPT-2-03 Rev.0
. \
l I SEGrrxJ/R-006, Trojan Second Integrated Test Data Report ! $$ liters of gas was collected in the bag and the calculated volume was 9.13 liters. The l Tedlar bag volume was used in calculating hydrogen content at 0.0574 g-mole H i . This run was above the acceptance value of 0.05 g-mole Hi . The sampling handling l blank correction of 0.007 g-moles H, was not subtracted from this result. l
! Figure 3.8 Run #2A Blamed Sample Analysis I
! i*
i j 1500
^A 1 1400
} d-. % ;
1200 - l E 1100 **
- 000
\_ "
~
g 7 , 700 600
-100 0 100 200 300 400 500 600 700 Run Time (min) 1 l
3.9 Run #2A Residue Random Sample - The quartz tube H2baseline evolution was at 769 ppm following an all-night baking at an Ar purge rate of 33 ml/ min with the furnace at i100*C. From -220 min to -4 min (in Figure 3.9), the H 2level decreased from 761 to 752 ppm. The baseline scatter was 2 4 ppm. This baseline was easily extrapolated for the duration of the run of 377 min. The residue was taken from the reprocessed Run #2A. It was uniformly sampled and j 11.90 g was then placed into a clean quartz boat. The tube was then opened, the boat was inserted into the hot zone, and the tube closed. Figure 3.9 show the H2data as a function of time from the start of run at 0 min. 'After 377 min the H 2level was changing at a rate of <2 ppm per minute and the run was stopped. 8.91 liters of gas was collected in the bag; however, the calculated volume of 12.44 liters was used in calculating hydrogen content at 0.1915 g-mole Hi . This run was well above the acceptance value of 0.05 g-mole Hi . The sampling handling blank correction of 0.007 g-moles H, was not subtracted from this result. l i i l TIO.RPT.2-03 Page 54 of 87
SEG/TRJ/R-006, Trojan Second Integrated Test Data Report M so=== samac .=w. .c. 1 Figure 3.9 Run #2A Random Sample Analysis , l 3000 A 2500 g 2000 \ h1500 z_ _ ..; 3000
; ; ; ^^ ; ;
-300 -200 100 0 100 200 300 400
{ l , Run Time (min) 3.10 Run #2A Residue 2nd Biased Sample The quartz tube H2 baseline evolution was at 677 ppm following an all-night baking at an Ar purge rate of 28.5 ml/ min with the fumace at 1100*C. From -34 min to -4 min (in Figure 3.10), the H 2level increased from 631 to 684 ppm, averaging 677 ppm. l This baseline was easily extrapolated for the duration of the run of 235 rrdn. . The residue was taken from the re-processed Run #2A. It was biasly sampled for the organic char residue rather than the metal dross. 12.70 g was then placed into a clean quanz boat. The tube was opened, the boat was inserted into the hot zone, and the tube closed. Figure 3.10 shows the H2 data as a function of time from the start of run at 0 min. After 235 min the H 2level was changing at a rate of <4 ppm per minute and the mn was stopped. 9.09 liters of gas was collected in the bag plus sample volume and the calculated volume was 6.56 liters. The Tedlar bag volume was used in calculating , hydrogen content at 0.1338 g-mole H i . This run was above the acceptance value of ) 0.05 g-mole H i. The sampling handling blank correction of 0.007 g-moles Hi was not subtracted from this result. This replicate run was biasly sampled to contain more organic char, but was pulverized to fine powder. It was felt that pulverizing the sample resulted in less ' scatter in the GC data, and possibly a higher hydrogen measurement due to the increased surface area and a more complete measurement of hydrogen in the residue l material. Based on these data all samples analyzed after 5/26 where ground prior to placing in the fumace. Previous samples may have analyzed low for H i , but since the . biased samples of previous runs did not meet the Hi limit, previous samples need not l be reanalyzed. TIU-RPT 2-03 Rev.0 . _ Page 55 of 87 l
SEGnRJ/R-006, Trojan Second Integrated Test Data Report @f@ so c imum .== .c. Figure 3.10 Run #2A 2nd Blased Sample Analysis 1700 <
' 5" 1
- w. _ . . . .
E! ,300,og tre r :f goo j 700 W' 500 50 0 50 100 150 200 250 ;
~
Run Time (min) 3.11 Run #3 Residue Blased Sample Run #3 contained large pieces of PVC and Raschig rings. The quartz tube H2 evolution averaged 1363 ppm following an all-night baking at an Ar purge rate of 14 m1/ min with the fumace at 1100*C. The flow through the glove box was the normal If/ min. Prior to the run the H 2level decreased from 1368 to 1349 ppm, or -
<4 ppm / min. This stable baseline was easily extrapolated for the duration of the run -
of 317 min. The residue was taken from Run #3. It was biasly sampled for the organic char residue rather than the metal dross; however, later review indicates that the sample was actually biased toward metal dross lumps rather than organic char in that white coated lumps were not specifically picked.14.3 g of residue was then placed into a clean quartz boat. The tube was then opened, the boat was inserted into the hot zone, and the tube closed. The tube Ar flow was 14 ml/ min. Figure 3.1I shows the H2 data as a function of time from the start of run at 0 min. After 317 min., the H level 2 was at 1873 ppm, and only changing about 6 ppm / min. Although this was not close to the extrapolated baseline of 1363 ppm, the run was completed. 5.57 liters of gas was collected in the bag compared to the calculated volume of 4.44 liters. The Tedlar bag volume was used in calculating hydrogen content at 0.1728 g-mole H,. The integrated level plus baseline was 2789 ppm, the Tedlar gas bag concentration was not available. This iun was above the acceptance value of 0.05 g-mole H,. The sampling handling blank correction of 0.007 g-moles - H, was not subtracted from this result. l TRJ.RPT-2 03 Rev.0 l Page 56 of 87 i
SEG/rRJ/R-006, Trojan Second Integrated Test Data Report @@ soom ecou. x,. .c Figure 3.11 Run #3 Blased Sample Analysis 5000 4000 E R Q 3000
'(
2000 3 _._ l 1000
-100 0 100 200 300 400 Run Time (min) 3.12 Run #3 Residue Random Sample
~
Run #3 ' contained large pieces of PVC and Raschig rings. The quartz tube H 2 evolution was at 1098 ppm following an all-night baking at an Ar purge rate of 42 ml/ min with the furnace at i100*C. The flow through the glove box was the normal 1 Umin. Prior to the run the H2 level was 978 ppm. From -70 min to -4 min > (Figure 3.12), the H2level decreased from i101 to 978 ppm (average 1098 ppm), or
<4 ppm / min. This stable baseline was easily extrapolated for the duration of the run i of 347 min. l The residue was taken from Run #3. It was randomly sampled for the organic char ,
residue and the metal dross.18.93 g was then placed into a clean quartz boat. The l tube was then opened, the boat was inserted into the hot zone, and the tube closed. l The tube Ar flow was 41.9 ml/ min. l Figure 3.12 shows the H2 data as a function of time from the start of run at 0 min. ' After 432 min., the H level 2 was at 1096 ppm, and only changing about 6 ppm / min. Since this was less than the extrapolated baseline of 1098 ppm, the run was completed. 15.15 liters of gas was collected in the bag compared to the calculated volume of 15.07 liters. The Tedlar bag volume was used in calculating' , hydrogen content at 1.4151 g-mole H i . This run was well above the acceptance value of 0.05 g-mole H i . The sampling handling blank correction of 0.007 g-moles Hi was not subtracted from this result. Figure 3.12 Run #3 Random Sample Analysis 40500 d l
! s oosoo g
h20500 10500
\
500-50 0 50 100 150 200 250 Run Time (min) l l TRJ.RFr.2 03 Rev.0
- Page 57 of 87 l
l
SEG/TRJ/R-006, Trojan Second Integrated Test Data Report $@@ 3.13 Run #3 Residue White Coated Lumps Run #3 contained large pieces of PVC and Raschig rings. The quartz tube H 2 . evolution was at 734 ppm following an all-night baking at an Ar purge rate of 42.6 ml/ min with the furnace at 1100*C. The flow through the glove box was the normal i Umin. From -51 min to -4 min (Figure 3.13), the H 2level increased from 724 to 746 ppm, or <4 ppm / min. This stable baseline was easily extrapolated for the duration of the run of 205 min. The residue was taken from Run #3. It was biasly sampled for the white coated residue lumps rather than the metal dross.19.45 g was then placed into a clean quartz boat. The tube was then opened, the boat was inserted into the hot zone, and the tube closed. The tube Ar flow as 42.6 ml/ min. Figure 3.13 shows the H2 data as a function of time from the start of run at 0 min. After 205 min, the H level 2 was at 2006 ppm, and only changing about 6 ppm / min. 8.86 liters of gas was collected in the bag compared to the calculated volume of 8.76 liters. The Tedlar bag volume was used in calculating hydrogen content at i l 1.5001 g-mole H i . This run was above the acceptance value of 0.05 g-mole Hi . 'Ihe l
- sampling handling blank correction of 0.007 g-moles H iwas not subtracted from this
- result. 1 L l l i Figure 3.13 Run #3 White Coated Lumps Analysis 40500 i
n 30500 J\ c.20500 10500 y .-. n aw+
-50 0 SO 100 150 200 250 Run Time (min) 3.14 Run #4 Residue Sample i This run was terminated prior to sampling owing to the piping heater failure in the CFE.
3.15 Run #5 Residue Biased Sample This mn had no PVC in the surrogate waste. The quartz tube H2 evolution was at 637 ppm following an all-night baking at an Ar purge rate of 31 ml/ min with the furnace at 1100*C. The flow through the glove box was the normal 1 Umin. From TRJ.RPT 2-03 Rev.O
SEG/fRJ/R-006, Trojan Second Integrated Test Data Report i M
-89 min to -4 min (Figure 3.14), the H2level varied from 649 to 620 ppm. This stable baseline was easily extrapolated for the duration of the run of 370 min.
The residue was taken from Run #5. It was biasly sampled for the organic char residue rather than the metal dross. 17.6 g was then placed into a clean quartz boat. The tube was then opened, the boat 'was inserted into the hot zone, and the tube closed. The tube Ar flow was 31 ml/ min. Figure 3.14 shows the H 2data as a function of time from the start of run at 0 min. After 370 min, the H level 2 was at 708 ppm, and only changing about <4 ppm / min. Since this was close to the extrapolated baseline of 637 ppm, the run was stopped. 9.62 liters of gas was collected in the bag plus samples compared to the calculated volume of 11.47 liters. The calculated volume was used in calculating hydrogen content at 0.0387 g-mole H .i This run met the. acceptance value of 0.05 g-mole H i. The sampling handling blank correction of 0.007 g-moles H iwas not subtracted from this result. U Figure 3.14 Run #5 Blased Sample Analysis 1200 1000' k . E s00 [ 1 h_ -~ l *
, 40o l I l l 1 -100 0 100 200 300 400
?I Run Tim e (min) l 3.16 Run #5 Residue Random Sample
)
This run had no PVC in the surrogate waste. The sample had been left exposed to the glove box atmosphere for over 24 hours. The quartz tube H2 evolution was at 559 ppm following an all-night baking at an Ar purge rate of 28 ml/ min with the furnace , at i 100*C The flow through the glove box was the normal f/ min. From -69 min to - l 4 min (Figure 3.15), the H 2level varied from 543 to 579 ppm. This stable baschne : was easily extrapolated for the duration of the run of 327 min. l The residue was taken from Run #5. It was uniformly sampled for the organic char l residue and metal dross.17.4 g was then placed into a clean quartz boat. The tube l was then opened, the boat was inserted into the hot zone, and the tube closed. The , tube At flow was 28 ml/ min. l TRJ RFT 2 03 Rev.0 Page 59 of 87 , 1
SEG/TRJ/R-006, Trojan Second Integrated Test Data Report @@
== == .m .c Figure 3.15 shows the H2 data as a function of time from the start of run at 0 min.
After 327 min, the H 1evel 2 was at 768 ppm, and only changing about <4 ppm / min. Since this was close to the extrapolated baseline of 559 ppm, the run was stopped. 8.3 liters of gas was collected in the bag plus samples compared to the calculated volume of 9.16 liters. The calculated volume was used in calculating hydrogen content at 0.0516 g-mole Hi . While the sample was left standing out in the glove box atmosphere,it may have absorbed some moisture. DreagerTube analysis of the glove box atmosphere showed approximately 500 ppm moisture. To determine the potential of a sample gaining moisture from exposure in the glove box, an already analyzed sample from Run #8 was exposed to glove box atmosphere for 6 hrs and reanalyzed at 0.005 g-mole H i. This run met the acceptance value of 0.05 g-mole H i when corrected for moisture contamination. The sampling handling blank correction of 0.007 g-moles H iwas not subtracted from this result. l I i i Figure 3.15 Run #5 Random Sample Analysis 1200
!! 200o i ! IAl l 1 Il E. 800 --
i el
" 8 400 e-l lM}
l l i i j' -300 -200 100 0 100 200 300 400 Run Tim e (min) 3.17 Run #6 Residue Blased Sample j This run had large chunks cf PVC added to the surrogate waste. The quartz tube H 2 evolution was at 810 ppm following an all-night baking at an Ar purge rate of 36 ml/ min with the fumace at 1100*C. The flow through the glove box was the normal 1 Umin. Prior to the run the H2 level was 809 ppm. From -100 min to -4 min I (Figure 3.16), the H2level remained steady at 809 ppm i 10 ppm. This stable baseline was easily extrapolated for the duration of the run of 64 min before it was terminated. The residue was taken from Run #6. It was biasly sampled for the organic char residue rather than the metal dross. 17.25 g was then placed into a clean quartz boat.
- The tube was then opened, the boat was inserted into the hot zone, and the tube j closed. The tube Ar flow was 36 ml/ min.
Figure 3.16 shows the H2 data as a function of time from the stan of run at 0 min. Initially the hydrogen peak soared to 88,527 ppm. After 64 min, the H2 level was at 45,581 ppm, and it was clear that the PVC chunks were not completely processed. l The run was terminated. Calculated hydrogen content based on the incomplete mn i TRLRPT 2-03 Rev.0
i
~
SEGiTRJ/R-006, Trojan Second Integrated Test Data Report $$@ N B00LDEF ew. oc l was 2.687 g-mole Hi . Estimating the remaining portion of the hydrogen curve would l suggest hydrogen levels as high as 6 g-me!es Hi . This was well above the acceptance value of 0.05 g-mole H i. l l \ j Figure 3.16 Run #6 Biased Sample Analysis l !! l i 100000 . l '! 80000
'A
' ! I E 60000 ch-[ e000 i
! l I
% ij 20000 : I il i ,
!~ 0 : r-kere 'reece ! I l
I ! -60 -40 -20 0 20 40 60 80
- j. Run Time (min) b {
- J l
3.18 Run #7 Residue Blased Sample l This run had large chunks of PVC added to the surrogate waste. The quartz tube H - l evolution was at 776 ppm following an all-night baking at an Ar purge rate of 38 ml/ min with the fumace at i100'C. The flow through the glove box was the normal i Umin. Prior to the run the average H 21evel was 776 ppm and was used for background. From -124 min to -4 min (Figure 3.17), the H 2level remained steady at 776 ppm 15 ppm. This stable baseline was easily extrapolated for 225 min before the mn was terminated. The residue was taken from Run #7, It was biasly sampled for the organic char, including white coated lumps, residue rather than the metal dross. 18.7 g was then placed into a clean quartz boat. The tube was then opened, the boat was inserted into the hot zone, and the tube closed. The tube Ar flow was 40.7 ml/ min. l Figure 3.17 shows the Hz data as a function of time from the start of run at 0 min. The
- residue hydrogen tests showed an evolution curve that had a very rapid and sharp
! increase to 44,678 ppm and then rapidly declined toward the initial background. After 225 min, the H level 2 was at 1537 ppm so the run was stopped, the residue generated 1.650 g-moles H i , and it was clear that the PVC chunks were not completely processed. The sampling handling blank correction of 0.007 g-moles H i was not subtracted from this result. Applying the dross correction of Section 3.24 yields 0.7348 g-moles H i. This was well above the acceptance value of 0.05 g-mole H.i TRJ.RFT 2 03 Page 61 of 87
SEG/TRJ/R-006, Trojan Second Integrated Test Data Report N@@
=== man .c.
!! Figure 3.17 Run #7 Blased Sample Analysis 40500 b l i l l l D 305o0 ;!
L.z i t, . i [20500 I Y I l : ( <l I
~
soo ' j r p
-50.00 0.00 50.00 g. 50.00 200.00 250.00 e
~
3.19 Run #7. Residue "No White" Sample , This run is a repeat of the above run, however, the sample was collected so that no white coated lumps were placed in the sample boat. Even after the sample had been crushed, any white material was removed prior to placing the powder into the quartz boat. The residue was taken from Run #7.17.8 g of material was selected, ground, ! and then placed into a clean quartz boat. The tube was then opened, the boat was inserted into the hot zone, and the tube closed. The tube Ar flow was 39 ml/ min. The quartz tube H2evolution averaged 2000 ppm, and was declining from the - previous run. The Ar flow rate was 39 ml/ min with the fumace at 1100*C. The flow in the glove box was the normal 1 t/ min. From -82 min to -4 min (in Figure 3.18), the H 2level dropped from 2552 to 1433 ppm. This baseline was easily extrapolated for the duration of the 256 minute run. Figure 3.18 shows the H2 data as a function of time from the start of the run at 0 min. This hydrogen curve was less than 1/20 of the previous analysis, indicating that the white coated lumps contained high levels of hydrogen. After 256 min, the H2 level was 878 ppm and changing <6 ppm / sample. The run was terminated, a calculated volume of 7.55 t of Ar was sampled, averaging 463 ppm. with a final concentration of 0.0659 g-moles Hi generated. The sampling handling bLnk correction of 0.007 g-moles H, was not subtracted from this result. Applying the dross correction of Section 3.24 yields 0.0358 g-moles H,. With the correction this sample, which was biased toward dross only, meets the acceptance value of 0.05 g-mole Hi . This shows that most of the unprocessed organics were in the white coated lumps. Page 62 of 87
l l 1 l SEG/TRJ/R-006, Trojan Second Integrated Test Data Report @@@ somssesanos, m,..c Figure 3.18 Run #7 "No White" Sample Analysis , t' 3000.00
--n 2500.00 j' l Q 20m.m .
d1 ! l ! I ! h 1500.00 l: ' **
.i 500.00 i 7 ,* l I
-100 50 0 50 100 150 200 250 300 g Run Time (min) l l l i 3.20 Run #8 Residue Biased Sample (cloth) l This run had no PVC in the surrogate waste but used I" Bert Saddles instead of the 1" l
l Raschig rings used on earlier runs. The quartz tube H2 evolution was at 867 ppm following an all-night baking at an Ar purge rate of 40 ml/ min with the fumace at l 1100*C. The flow through the glove box was the normal 1 t/ min. From -85 to -4 ; l min (in Figure 3.19), the H 2level was steady at 867 10 ppm. This stable baseline l was easily extrapolated for the duration of the run of 277 min. The residue was taken from Run #8. It was biasly sampled for the unprocessed center - portions of organic char residue lumps and a piece of charred cloth that had l mistakenly been dropped or left in the process can. 19.06 g was ther. pt: red into a clean quartz boat. The tube was then opened, the boat was inserted into the hot zone, l and the tube closed. The tube Ar flow was 40 ml/ min. l Figure 3.19 shows the H2 data as a function of time from the stan of run at 0 min.
- After 277 min the H level 2 was at 927 ppm, and only changing about <4 ppm / min.
! Since this was close to the extrapolated baseline of 867 ppm, the run was stopped. l 10.99 liters of gas was collected in the bag compared to the calculated volume of i 11.08 liters. The calculated volume was used in calculating hydrogen content at O.0811 g-mole Hi . The sampling handling blank correction of 0.007 g-moles H, wr: , not subtracted from this result. This was above the acceptance value of 0.05 g-mole H,. 1 l l Figure 3.19 Run #8 Blased (Cloth) Sample Analysis l j 2500.00 i ' l I I lN I ! I l l~ 2000.00 j 1 ; i
!, 1500 00 , .
j i
', " teco co / , i c, , ,, m ],
, 500 00 i!
l 100 50 0 50 100 150 200 250 300 *' Run Tim e (min) h TRJ.RPT 2 03 Rev.0 i
SEG/TRJ/R-006, Trojan Second Integrated Test Data Report $@@
==== ==u. -@.c 3.21 Run #8 Residue Biased Duplicate Sample ("No Cloth")
This run had no PVC in the surrogate waste. The quartz tube H2evolution was at 673 ppm following an all-night baking at an Ar purge rate of 40.5 ml/ min with the furnace at 1100*C, The flow through the glove box was the normal 1 t/ min. From -33 min to '
-4 min (Figure 3.20), the H 2level varied from 666 to 679 ppm. This stable baseline was easily extrapolated for the duration of the mn of 345 min.
The residue was taken from Run #8. It was biasly sampled for the unprocessed center portions of organic char residue lumps and no pieces of cloth.19.35 g was then placed into a clean quartz boat. The tube was then opened, the boat was inserted into the hot zone, and the tube closed. The tube Ar flow was 40.5 ml/ min. Figure 3.20 shows the H2data as a function of time from the start of run at 0 min. After 345 min, the H level2 was at 699 ppm, and only changing about <4 ppm / min, Since this was close to the extrapolated baseline of 673 ppm, the mn was stopped. 13.22 liters of gas was collected in the bag compared to the calculated volume of 13.97 liters. The calculated volume was used in calculating hydrogen content at 0.0292 g-mole H i. The sampling handling blank correction of 0.007 g-moles Hi was not subtracted from this result. This run met the acceptance value of 0.05 g-mole H i. Apparently, the cloth residue from a cleaning rag was the cause of the high hydrogen content measured in Section 3.20. Figure 3.20 Run #8 Biased "No Cloth" Sample Analysis 1000 lk I I ! ,
,o 16 % i I i i a00 I I 700 , - --
600 7 I I I
-100 0 100 200 300 400 !
l Run Time (min) 3.22 Run #9 Residue Blased Sample of 1% of 0.25" PVC at 100*F Hotter
, l This run experimented with smaller sized PVC pieces of 0.25 inches. This mn also utilized a 100*F hotter temperature at the process can inlet and 56*F hotter
! temperature at the process can outlet. The temperature increases and smaller PVC particle size proved important. ; l 1 This run had only 53.5 grams of PVC or 1% of the total sample in the surrogate waste. The quartz tube H 2evolution was at 765 ppm following an all-night baking at an Ar purge rate of 43 ml/ min with the fumace at 1100*C. The flow through the glove box was the normal if/ min. From -43 min to -4 min (Figure 3.21), the H2 level TRJ.RPT-2 03 Rev.O p
SEG/TRJ/R-006, Trojan Second Integrated Test Data Report
= =@@
l
= .m@ 1 l
varied from 773 to 755 ppm. This stable baseline was easily extrapolated for the duration of the run of 304 min. The residue waste taken from Run #9 was biasly sampled for the unprocessed center portions of organic char residue lumps,18.50 g was then placed into a clean quartz boat. The tube was then opened, the boat was inserted into the hot zone, and the tube I closed. The tube Ar flow was 43 ml/ min. Figure 3.21 shows the H2data as a function of time from the start of run at 0 min. After 304 min, the H level 2 was at 899 ppm, and only changing about <4 ppm / min. Since this was close to the extrapolated baseline of 765 ppm, the mn was stopped. 14.35 liters of gas was collected in the bag compared to the calculated volume of 13.07 liters. The Tedlar bag volume was used in calculating hydrogen content at 0.0979 g-moles H i . The sample handling blank correction of 0.007 g-moles Hi was not subtracted from this result. Applying the dross correction of Section 3.24 yields 0.0302 g-moles H i. The corrected value met the acceptance value of 0.05 g-mole Hi . l I '- 1 Figure 3.21 Run #9 Biased Sample Analysis 3 l l! --
;; 2150 N-i 1950 'OM i ' '
'N 1750 I % I Z ' % ' ' '
1550 - 1350 'O E' ' 1150 : 'k ' ' 950 I N - .. . -- . . . . ! , 750 y-5 .
-100 0 100 200 300 400 n Run Time (min) g 3.23 Run #9 Residue Blended Random Sample of 1% of 0.25" PVC at 100*F Hotter For this run, half of the Run #9 residue was placed into a mortar, and was pulverized with a pestle. The ground residue was then filtered through a sieve. Any pieces remaining in the sieve were reground. The material was powder prior to any random l sampling occurring. The residue was randomly sampled from the powdered material l
and 20.22 g was then placed into a clean quartz boat. The tube was then opened, the boat was inserted into the hot zone, and the tube closed. The tube Ar flow was 44.4 ml/ min. l The quartz tube Hz evolution was at 1024 ppm following an all-night baking at an Ar ! purge rate of 44.4 ml/ min with the fumace at 1100*C. The flow through the glove box was the normal 1f/ min. From -43 min to -4 min (Figure 3.22), the H2 level varied from 1003 to 1034 ppm. This stable baseline was easily extrapolated for the duration of the run of 331 min. l TRJ-RFT 2 01 Page 65 of 87
1 SEG/TRJ/R-006, Trojan Second Integrated Test Data Report @@
==== === =@.c Figure 3.22 shows the H2 data as a function of time from the start of run at 0 min.
After 330 min, the H level 2 was at 1091 ppm, and only changing about <4 ppm / min. 1 Since this was close to the extrapolated baseline of 1024 ppm, the run was stopped. The volume of gas collected in the Tedlar bag was not available as it did not get ; connected prior to commencing the run. In all runs analyzed after Jun 10, the , difference between Tedlar bag and calculated volumes was less than 10%. Therefore, i the lack of a Tedlar bag volume would indicate a possible 10% error introduced in the results. This error was not enough to affect the final results. The calculated volume was 14.70 liters. The result was 404 ppm, or 0.0923 g-moles H . The sample 3 handling blank correction of 0.007 g-moles H was not subtracted from this result. Applying the dross correction of Section 3.24 yields 0.0285 g-moles Hi . The corrected value met the acceptance value of 0.05 g-mole Hi . l Figure 3.22 Run #9 Random Sample Analysis 1 2700 - I
! 2e ! ' I ' ' i l I l
$2100 Nk I ' '
l
% l I l
h[1800 a 1500 iY ' 12m i % i l [i . .00 e Runhimekin) , l i 1 3.24 Residue Hydrogen Content Computation (Dross Correction)
, In accordance with the Integrated Test Sampling and Analysis Plan (SEGfrRJIPRO-032), the H 2concentration remaining in the residue after steam-reforming the waste sample was calculated from the mass of the residue taken out of the process can, the measured H2 level collected in the gas bag, the volume of Ar-H 2in the gas bag, the density of the gas in the sample bag, the mass of the residue sample placed into the quartz furnace boat, and the molecular weight of mono-atomic hydrogen, H ,i as follows:
G-moles Hi = 2.(e residue m WH, in bag. ppmWBag Vol. L 1.(o. gH,Lal in residue 1x108.(2.01594 gH2 /g-moleH 2)*(g boat sample) where: L = liters of gas, and p = gas density calculated by: p = P = Mw/R.T (typically around 0.083 gHA). Where: P = 1 atmosphere Mw = molecular weight of H2= 2.01594 TRJ.RPT-2-03 Rev.0
SEG/TRJ/R-006, Trojan Second Integrated Test Data Report t )@@@
== = umoa ..c R = Universal Gas Constant = 0.08205 L = atmos /g-mole ="K T = Gas Bag Absolute Temperature
- K (approximately 294.82*K)
A summary of the calculated mono-atomic hydrogen, H ,i results for each run is presented in Table 3.1. The calculation of g-moles H assumed i that the dross portion of the residue contributes equally with the surrogate waste to the overall hydrogen content in a process can. In fact, no significant amount of hydrogen is contributed by the dross because it is pure stainless steel powder; therefore, the method of calculating total hydrogen (H )3 in a process can requires revision to be more representative. The dross weight must be subtracted out of the calculations defined above and in the Trojan Integrated Test Sampling and Analysis Plan (SEGfTRJIPRO-032). s The total grams of dross added to the process can before processing must be subtracted from the parameter "g residuca ," in the numerator. In addition, the grams of dross in the sample boat must be subtracted from the parameter "g boat sample" in the denominator. To accomplish this, two correction ratios must be developed. Numbers used are from Run #9 biased sample analysis. A. Numerator correction. Mass of total residue used to calculate the total H content of 0.087 g-moles in the process can was 3852 grams. The dross added to the process can was 3635.6 grams. Therefore, the correction ratio is : 3852-3635.6/3852 = 216.4/3852 B. Denominator correction. Mass of sample collected for the boat used to calculate the total H, content of 0.087 g moles in the process can was 18.5 grams. The visual ratio of PVC lumps to dross lumps in the boat was 40 volume percent white coated lumps which contain PVC residue and 60 volume percent darker lumps which contain dross residue. Since the blended and biased samples for Run #9 calculated to nearly the same (0.085 and 0.087 respectively), the density difference between PVC residue and dross residue is probably minimal. For this calculation, the density ratio of dross residue to PVC residue is estimated very conservatively at 3 to 1. Therefore, the weight ratio of PVC to dross residue is: (0.40/0.60)(0.25/0.75) = 1 g PVC residue / 4.5 g dross residue 18.5(1/5.5) = 3.4 grams of PVC residue in the boat To correct the value of 0.087, apply the above corrections: 0.0979[(18.5/3.4)(216.4/3852)] = 0.030 g moles H, in the process can Five process cans per process can capsule results in 0.150 g moles H, per capsule. Page 67 of 87
i SEG/TRJ/R-006, Trojan Second Integrated Test Data Report u )@@ a.= i= .c The dross correction was applied to Runs #7 and #9 and recorded in Table 3.1. Insufficient data was available to apply to earlier runs. 3.25 Data Precision, Accuracy, Completeness, Comparability, and Percent Relative Standard Deviation 3.25.1 Data Precision Precision is an indicator of the repeatability of a measurement, and is expressed as the relative percent difference (RPD) of two analytical measurements. The formula for calculating the RPD of two measurements is shown below. RPD= (SA . (S+D)/2 Where: S= The larger anakysis result of two measurements D= The smaller analysis result of two measurements Almost all of the Runs were analyzed twice, once was a bias sample and the second was a random sample. Unfortunately, because of these differences the majority of analysis can not be used to determine data precision. Two Runs, #2 and #3 were analyzed three times each. Run #2 was biasly sr.mpled twice, and Run #3 was randomly sampled twice (one was completely random, the other was a visually representative sample). Putting the data from Runs #2 and #3 into the equation provided above we obtain RPD values of : 45.6% for Run #2, and 5.2% for Run #3. The difference in the RPD values is most likely due to the sampling methods used. It is very difficult to produce identically biased samples, however, if a sample is tmly representative (or random) the analysis should be repeatable, as shown by the data collected from Run #3. The data quality objective of 67 percent for RPD was met. 3.25.2 Data Accuracy Data accuracy is an indication of how close a measurement is to the actual value. Data accuracy is determined by spiking a sample with a known value and measuring the percent recovery (%R) of the spike. The formula for calculating data accuracy is shown below.
%R= 1004 R-SM SP Page 68 of 87 L
j SEG/TRJ/R-006, Trojan Second Integrated Test Data Report
==-c e
@@ .c Where: SSR = Spike sample result (H in mg)
SR = Unspiked sample result (H in mg) SP = Amount of spike added (H in mg) The percent recovery was measured for three separate phenanthrene spikes. Using the data obtained from the phenanthrene runs, and the blank run, and l putting it into the equation provided above we obtain the following %R: Phenanthrene Run #1 78 % Phenanthrene Run #2 64 % j I Phenanthrene Run #3 244 % l Runs #1 and #2 %R are within the acceptable limits of 60 to 140 percent. It is ! hypothesized the drastic increased recovery in Phenanthrene Run #3 is due't'o : the sample boat being inserted too far into the reactor tube, " jostling" the carbon I l bed. This would explain the unusual increase in hydrogen background which l occurred during Phenanthrene Run #3. i l 1 3.25.3 Percent Relative Standard Deviation l The % RSD will be applied to the instrument response factors for standard l analyses as a mesure oflinearity of the GC instrument calibrations. The l
- % RSD is calculated using the following formula
- i t
i i
%RSDn STDEV"^
R, Where: STDEVg = Sample population standard deviation R m = Average of analytical results - The GC was calibrated with a 6-point calibration curve. The standards used i included 10,50,100,200,500, and 3000 ppm standards. Ideally, the %RSD for l the hydrogen calibration would be less than 20%. However, as the calibration l curve gets closer to the detection limit of the instrument, the %RSD may increase. The GC was calibrated several times during the Integrated Test, providing the following data: Calibration Correlation %RSD
#1 5/8/97 .99959 15.38
#2 5/12/97 .99992 25.10
#3 5/24/97 .99992 90.16
#4 5/28/97 .99996 99.62
#5 6/17/97 .99999 38.74 TRJ-RFr.2-03 Rev.0 l Page 69 of 87 l
l SEG/TRJ/R-006, Trojan Second Integrated Test Data Report $@@
=== me. , .c j
' A standard check was completed every morning, any time the standards were more I than 10% off the expected value the GC was recalibrated. The initial calibration was l completed, and all 6 standards were detectable. A sample of 2.31 grams of :
phenanthrene was erroneously run through the reactor tube and GC, after which we ; were no longer able to detect the 10 ppm standards. This left us with a five point calibration for #2, #3, and #4. Throughout the Integrated Test we saw the ability to ) detect the 50 ppm standards decreasing, this affected the %RSD. By the time l calibration #5 was completed the 50 ppm standards were no longer detectable, and j were not included in the calibration, thereby drastically decreasing the %RSD.
]
3.26 Thermogravimetric /GC Analyses at IT Lab The importance of the process can processing temperature was explored employing an analytical technique called Thermogravimetric / gas chromatographic analysis (TGA/GC). This method allows a sample to be pyrolyzed with Ar inert gas as a function of small temperature steps from 600 to 1100*C at the same time the gases ; evolved are analyzed by chromatography. A total weight loss is determined at the ! conclusion of pyrolysis. I In Figure 3.23 shows the gas evolution of H 2and H O 2 as a function of temperature for a sample of the PVC. The raw PVC was reduced by 85% in mass. The largest amount of gas evolved was in the temperature range from 400 - 600*C. Note that the area under each of these gas peaks following the 400 - 600*C peak is smaller and decreasing in size as the temperature increases from 600 - 1100*C. These results show that by 1100*C no significant gas is evolved - furtherjustifying the selection of 1100'C for the quartz tube furnace. In addition, the total quantity of hydrogen evolved from the raw PVC plastic that was not steam reformed was about 20 g-moles - further supporting the 99.6% (20/0.05) reduction of the hydrogen containing substances during the steam-reforming process. The condensate that collected in the apparatus was pyrolyzed and analyzed by the GC at the end of the thermal analysis to include the hydrogen evolved in the final calculations. This condensate formed the final peak in Figure 3.23. l i ! Figure 3.24 shows the gas evolution of H2and H O 2 as a function of temperature for a sample of the steam reformed residue from Run #7 that was hand-selected for white coated high organic bearing lumps. The nature of these white coated lumps has been L discussed earlier. These lumps were produced in the normal steam-reforming of the waste in Run #7, and these lumps were subjected to TGA/GC analysis. 4 Because of the large variation in sample size and argon flow rates between the l analyses illustrated in Figure 3.23 through 3.27, the percent of total for each peak . { must be compared rather than absolute values. Note, that in Figure 3.24 the area 1 under the peaks at the lower temperature intervals from 400-600*C were reduced by the CFE steam-reforming chemistry at 500-600*C from those from raw PVC as l TRJ.RPT.2-03 Rev.0 l
m . . _ _ _ _ _ . _. ._ __ _ _ _ _ _ . _ . _ _ _ _ . _ _ _ l
- SEG/TRJ/R-006, Trojan Second Integrated Test Data Report LM@@
=== unu. .c shown in Figure 3.23, but the upper high temperature peaks were not affected. These high temperature intervals evolved about the same amount of hydrogen as the raw PVC in this temperature range. Thus, this effect validates the hypothesis that the l-white coated lumps are associated with incompletely processed PVC/ crude oil.
Figure 3.25 shows the TGA/GC results for a fine ground (~1/32 inch particle size) sample of raw PVC. First, the raw sample was steam reformed for 320 min at 600*C. l Then the material was dried in Ar flow for 124 min. Finally the steam-reformed and dried residue was hydrogen tested by stepping the temperature steadily to 1100*C. The evolution of H O 2 and H is 2 plotted as the temperature is stepped to 1100*C over i 219 min. First, it is noted that 99.94% of the sample was decomposed by steam-reforming and j drying over 484 min, and only 0.056% of the sample decomposed above 600'C as shown by the gradual rise in H2 O and H2 evolution curves over this last region. The high temperature H2peak was only 50 ppm, which would be in the digital noise of the H 2sensor background in the Integrated Test. l These are very significant results that show that PVC reduced to 1/32 inch size steam-reforms almost quantitatively with residual hydrogen evolution at or near level of detection and less than the boat blank. Thus, whatever fraction of PVC that is around 1/32 inch in size will not contribute significant hydrogen in the steam reformed residue. ! l l Figure 3.26 shows the gas evolution of H2 O and H2 as a function of temperature for a sample of actual radioactive PVC taken fmm the Trojan Spent Fuel Pool. The sample was analyzed using the same methodology for the raw PVC shown in Figure 3.23. As , in Figure 3.23, the last peak represents the pyrolysis of condensate at the end of the j thermal analysis. The close similarity of Figure 3.26 with Figure 3.23 suggests that the Trojan sample of actual waste is made of the identical material as the raw PVC used as surrogate waste in the Integrated Tests. Figure 3.27 shows :he gas evolution of H2 O and H2 as a function of temperature for a sample of PVC, containing 80% ground PVC and 20% pieces of PVC about 1/8 to ! 1/4 inches in diameter, which was steam reformed and then subjected to TGA/GC l analysis. When compared to Figure 3.24, the reduced size of the higher temperature l peaks shows that steam reforming efficiency of PVC improves with decreasing ; i particle size. i l l 1 b l TRJ-RFT-2-03 Rev.O Page 71 of 87 4
l l SEG/TRJ/R-006, Trojnn 2nd Integret:d Test Data Rep rt j FIGURE 3.23
SUMMARY
OF GAS EVOLUTION AS A FUNCTION OF TEMPERATURE FOR RAW PLASTIC SAMPLE 5000
-+- H2O *
--o- H2
{3500.
.S 3000. i -
' i d 2500 ,
! I 2000. ! i
! ! 9 I 1500 ' i l I i ! ! l O
1000 Ia ! J l bi 'Q2 9 500 _ _ . .) _ n-l 0 g.: N Y Y
- h hk j
7 - R , ~ $ $ l o ' Temperature (Deg. C) Sample Wt. = 0.1997 g Wt. loss = 0.1625 g H2 w!o w/Conden. H2O w/o Conden. w/Conden. Conden. T Range G-Moles H % of Total % of Total G-Moles H % of Total % of Total Evolved Evolved l 200C 8.74E-05 9.7 8.3 8.37E-05 9.3 8.0 l 200-400C 6.77E-05 7.5 6.4 4.94E-05 5.5 4.7 ! 400-600C 2.44E-04 26.9 23.2 1.10E-04 12.2 10.5 600-700C 1.58E-04 17.4 15.0 7.40E-05 8.2 7.0 l 700-800C 1.28E-04 14.2 12.2 6.62E-05 7.3 6.3 800-900C 9.39E-05 10.4 8.9 5.25E-05 5.8 5.0 900-1000C 8.05E-05 8.9 7.7 4.81 E-05 5.3 4.6 1000-1100C 4.53E-05 5.0 4.3 2.87E-05 3.2 2.7 Condensate 1.47E-04 14.0 5.42E-05 5.2 Total w/o 9.05E-04 5.13E-04 Condensate l Total w/ 1.05E-03 100.0 100.0 5.67E-04 56.7 53.9 ( Condensate
" Condensate remained in end of tube at completion of run. **No H2S detected in evolved cases l 1
TRJ-RPT-2-03 Rev.0
l l SEGfrRJ/R-006, Trcjnu 2nd Int:grat:d Tcst D2t2 R:psrt i l FIGURF.3.24
SUMMARY
OF GAS EVOLUTION DATA AS A FUNCTION l OF TEMPERATURE FOR REFORMED SAMPLE l 35000 4 30000 ' 1 l
- l. 2s.0 __m ,
_ _R ,
,' l j
i 20000 e H2O < L. I _7' 4<i g ,, #n "i<,. i i< - i ! l 0 # p2 ! l ,
$ 10000 t._._ .
5000 i '! 'l I l l
! T- - --
' ---' '- - -- l
, ...2-- --
0 i h 0 ! !e i O h O $h$b! - w r- r- m Temperature (deg. C)
- Sample Wt. = 4.2 g T Range G-Moles H Evolved % of Total 200-400*C 5.763E-06 0.03 400-600*C 7.031E-04 3.97 600-650*C 1.222E-03 6.90 650-700*C 2.524E-03 14.24 700-750*C 2.884E-03 16.28 750-800*C 2.706E-03 15.27 800-850*C 2.136E-03 12.06 1
850-900'C 1.201 E-03 6.78 900-950*C 1.376E-03 7.77 l l 950-1000*C 1.066E-03 6.02 i 1000-1050'C 9.584E-04 5.41 1050-1100*C 9.343 E-04 5.27 Total 1.77E-02 100.00 TRJ-RPT-2-03 Rev.0
( SEG/TRJ/R-006, Trajen 2nd Int: grated Tcet Dats Report ! I l FIGURE 3.25
SUMMARY
OF PYROLYSIS GAS EVOLUTION FOR GROUND STEAM REFORMED PLASTIC CHAR l r l 250 i l 1 ! l l I I
! i i !
% :: ; :: :::: e !?b ?rgee r:
, j ; I 1 !
l' 150 i ' ' ' ! I' ! H2O u 100 + 50 l 1 )
- - ~ ~ -" """" ' '-
0 -- W 8 a
~ $ a ;
s Temperature (Dog.C) l l l l Sample Wt. !nitial(g): 2.3448 Sample Wt. Final (g): 1.9700 84.02 % Argon Flow (ml/ min.): 41.4 Sample Wt. Loss (g): 0.3748 15.98 % i From Hydrogen Evolution J Run Time Total Avg.H2 l (min.) Volume Conc. G-Moles G-Moles H Percent of (L) (ppm) H Evolved Evolved per Total aram Steam Reform @ 600*C 118 4.89 27439.75 0.010931 0.004662 76.76552 i 230 4.64 3547.1 0.001341 0.000572 9.415007 320 3.73 2877.033 0.000874 0.000373 6.135558 Argon drying @ 600*C 484 7.76 1714.725 0.001086 0.000463 7.628207 ! Argon @ 600 - 1100*C 703 9.06 10.725 7.93E-06 0.000003 0.055705 Total 0.01424 0.006073 100 ; i l . TRJ-RPT-2-03 Rev.O
SEG/TRJ/R-006, Trojan 2nd Int: grated Test D:ta R: pert FIGURE 3.26 '
SUMMARY
OF GAS EVOLUTION DATA AS A FUNCTION OF TEMPERATURE FOR TROJAN PLASTIC SAMPLE 5000 I I I ' ' I I I i I I ! I 4500 ' i i l - 4000 id l ! ! ! l I t ! i i ! i l l
' I I ! '
! I ! ! I l ! ! I
{3500 i l i i !
-+ H2O I
l l l [ J 3000 I i I l ! ;, I i
! i ! ki I i i l i I I l 'T I -* H2 g2500 '
I ' oN I I I 'l N N *1 l 3 15m
* ':: u
]
- tUkA L I
A ikI D'*9 ; o n#IIN I ! ' I I I I I I l n E r- l E I b= Temperature (deg. C) Sample WtT. g): 0.2215
- Final Wt. (g): 0.0162 7.3%
Wt. Loss (at 0.2053 92.7 % H2 H2O T Range G-Moles H Evolved % of Total G-Moles H Evolved % of Total 200C 1.02E-05 1.1 6.92E-05 9.2 l
- 200-400C 6.48E-05 7.2 6.43E-05 8.5
! 400400C 2.42E-04 27.0 1.42E-04 18.9 S00-700C 1.11 E-04 12.4 8.89E-05 11.8 700-800C 1.07E-04 12.0 8.83E-05 11.7 800-900C 9.73E-05 10.9 9.07E-05 12.1 900-1100C 2.63E-04 29.4 2.09E-04 27.7 Total 8.96E-04 100.00 7.52E-04 100.00 ) PCondensate remained in end of tube at completion of run. TRJ RPT-2-03 Rev.0
SEG/TRJ/R-006, Trojen 2nd Integrcted Test Data Rspart FIGURE 3.27
SUMMARY
OF GAS EVOLUTION AS A FUNCTION OF TEMPERATURE FOR 80% GROUND AND 20% PIECES OF REFORMED PLASTIC 70000 60000 i ' ' ' ' '
bI !i '
! I I '! !
[ 00000 .M! I i i'! ' ' I i i i i ! [i
' I ! Ii'Il S 4o000 l l!I i Il!III l l 'll + "20 d 1 --e- H2
' I 8 30000 ' "
o, Tg,j '{ g l l WWg ". a H aha n l e l s Temperature (deg. C) Sample Wt. (g): 4.1906 : Final Wt. (g) 3.8576 92.1 % Wt. Loss (a): 0.3330 7.9% H2 H2O T Range G-Moles H Evolved % of Total G-Moles H Evolved % of Total 200C 1.46E-07 0.0 6.86E-05 2.4 200-400C 2.13E-05 0.1 8.23E-05 2.9 400-600C 3.75E-03 25.9 7.39E-04 25.7 600-700C 2.82E-03 19.5 4.48E-04 15.6 700-800C 3.32E-03 22.9 4.11 E-04 14.3 800-900C 2.18E-03 15.1 3.87E-04 13.5 900-1000C 1.25E-03 8.7 2.82E-04 9.8 1000-1100C 1.13E-03 7.8 4.56E-04 15.9 Total 1.45E-02 100.00 2.87E-03 100 00 TRJ-RPT-2-03 Rev.O
SEG/TRJ/R-006, Trojan Second Integrated Test Data Report _ h]@
== u- - .c
4.0 CONCLUSION
S AND RECOMMENDATIONS This section examines Integrated Test runs as well as the residue hydrogen content determinations and draws suggestions for running the Trojan Plant operation. 4.1 Operational Comments Throughout all test runs, the process can upper filter assembly showed little sign of plugging. The pressure drop across this filter has been monitored during the testing to show the small extent of increased pressure drop as particulate material gradually is filtered out of the gas flow. However, reusing process cans requires thorough cleaning of both the bottom and the upper filters to avoid low flow problems. Runs
#5 and #7 had the lowest steam flow rates through the process can, both runs were conducted with used process cans. Run #8 had normal steam flow with a used can, because the process can filters were cleaned with oven cleaner and a high pressure water sprayer. At Trojan, process cans will not be reused after processing actual spent fuel pool debris.
The process can/can feed evaporator bottom seal performed adequately for the early runs and runs after #6. For Runs #3 through #5, the flow control ring in the CFE l steam entry area restricted the seating of the can. New plugs were installed and the process cans seated satisfactorily for Runs #6 through #9. This seating was important , to ensuring that steam flow directed to the process can interior did not bypass around I the process can through the steam chamber. l Planned operating temperatures, flows and pressurcs were achieved per the data l quality objectives with few exceptions. Higher than expected pressure drops across the cesium trap resulted in more negative pressures at the SSR inlet, PT-107 than ! planned. These more negative pressures required higher blower speed, but did not ! hinder the SSR operations. Also noted was an increase in cesium trap pressure drop with successive runs. Replacing the cesium trap may be required more often than anticipated. A spare cesium trap has been procured and long lead materials have been procured for fabricating a third trap if required during the Trojan project. For all runs, heat up procedures have been designed to ensure that steam temperature at TE-K101 always equals or exceeds CFE temperature, TE-K102 until full process temperatures are achieved. This method of temperature control will prevent creating hard-packed melted masses in the waste which greatly affected flow rates and processing efficiency in Runs #2 and #4. l The cause of the melt down of Run #4 waste and termination of the mn was that the initially installed band heaters in the CFE steam supply piping failed during Run #3. In addition, the band heaters on the CFE and cesium trap loosened and slid down to the bottom of the outer chambers. Water dripping from the process can and subsequent condensate collected in the unheated steam inlet piping, blocking steam flow to the process can. The CFE was allowed to heat up without steam flow, nrm 03 Rev.0 l Page 77 of 87 - I
l l SEG/TRJ/R-006, Trojan Second Integrated Test Data Report )@@ l resulting in the hardened mass, no steam flow, and run termination. After Run #4, radiant heaters were installed in place of the band heaters on the piping and bracing l was installed for the CFE and cesium trap band heaters. In addition, element l temperature thermocouples and controllers were installed to prevent overheating. An l additional condensate drain was installed at the time the heat tracing was repaired to allow for the initial water collecting in the steam piping at the entrance to the process l can to be drained and reused. The steam superheater functioned flawlessly throughout the testing. A steam booster heater was installed prior to Run #9 to operate at elevated temperatures. Upon initial inspection, concern for possible outside air leaks into the CFE slide gate assembly stimulated the addition of a N2 Purge into the slide chamber area. This positive pressure purge countered any possible air leakage. Oxygen during most runs was typically below 0.5% After the leaking grapple device was repaired, a heated N2 Purge flow through the transfer bell kept this system dry even after dunking the grapple and chain in water. The water moistured measure by MS A tubes was 10 to 33 ppm. To provide a more direct reading of hydrogen content in the syn-gas, a hydrogen gas analyzer has been added to the gas analysis equipment in place of the CO2 analyzer used during the Proof Of Principal testing. This hydrogen sensor performed very well after the initial calibration and shakedown. The sensor showed peaks as high as 20,000 ppm and steady declines to its detection limit of140 ppm. The benzene sensor, added during Run #6, also performed well down to its detection limit of10.3 ppm. Stainless steel Raschig rings were placed in the process cans with the waste for Runs
#1 through #4. Runs #1 and #3 were very rapidly processed; however, biased samples taken from residue found inside of the Raschig rings found unprocessed organic material shielded from processing by the rings. Runs #5 through #7 did not include the Raschig rings. These runs took much longer to process. Runs #8 and #9 had stainless steel Berl saddles placed in the process cans with the waste. Both runs processed more rapidly and were successful in meeting the residual hydrogen criteria partly because there were no sheltered " rings" to shield the residue from processing.
The stainless steel Beri saddles will be included in process cans filled with Trojan spent fuel pool debris. During residue sampling for residual hydrogen analysis, biased and random sampling l techniques were employed. In runs where obvious unprocessed organics remained, ! Runs #1, and #7, the biased samples (biased toward the organic pieces) resulted in higher hydrogen content as expected. In runs which were completely processed, Runs
#5 and #9, random and biased sampling yielded nearly the same results. For Run #5, the random sample measured higher than the biased. This was due, at least in part, to the sample's exposure to moisture in the glove box atmosphere. For Runs #2 and #3, the biasing may have inadvertenly favored dross lumps rather than unprocessed l
TRJ Rirr-2 03 Rev.0
l l SEG/TRJ/R-006, Trojan Second Integrated Test Data Report . a,= .=u M[@ = .c organics. A second biased sample each for Runs #2 and #3 taken at a later date yielded results much closer to the random sample. Both the rerun biased samples and random samples for Runs #2 and #3 yielded nearly the same results. Since these were visually the least processed runs, the organics may have been so mixed in the residue that sampling technique made little difference. Runs #6 and #8 did not have random sample results available. Runs #5, #8 (no cloth), and #9 (with dross corrections) met the hydrogen criteria of l 0.05 g-moles H iper process can. The operating setpoints used to conduct Run #9 have been entered into the operating procedures for use at Trojan because all Trojan wastes may contain some PVC and because the values are more conservative, therefore, acceptable to apply for both PVC and non-PVC runs. Run #9 proved to be , successful in reforming 53.5 g of PVC cut into 1/4" sized pieces. I 4.2 In-line Hydrogen, Carbon Monoxide and Benzene in End-of-Run Determinations l The goal was to demonstrate the use of a variety of chemical indicating sensors that could be used to assist in the determination of end-of-run. All three, hydrogen, CO and benzene were useful. The most useful was hydrogen because of it higher sensitivity and ability to detect down to its detection limit of 40 ppm. Benzene was 10.3 ppm but the benzene dropped to below detection limit before the hydrogen, therefore, was not as useful by itself. All three sensors will be used at Trojan. As a backup, several spare hydrogen sensors, electronics, and calibration gas cylinders i have been ordered. End-of-run criteria to be used at Trojan has been determined to be s200 ppm H 2for three consecutive hours with CO and benzene at their respective detection limits. This criteria was used successfully in Runs #8 and #9. 4.3 The Challenge of " Neoprene" (sic PVC) When it became evident that the " neoprene," black flange filter ends would have to be processed, more inquiries about the composition details of the " neoprene" were made of the filter supplier. The surprise was that the " neoprene" that was used in the black flanges for the cartridge filters was actually plasticized PVC resin (containing ~50 wt% chlorine) that included the raw resin mixed with phthalic ester, polyols, zinc ! stearate " mold release" additive, calcium carbonate, carbon black colorant (~0.5%) and " crude oil" extender (about 30 wt %). The cartridge filter supplier was unwilling to supply proprietary compositions. But actual telephone conversations with Youtis Anathan, chief chemist at Diversified Compounders,Inc., supplier of the black i flanges for the cartidge filters, disclosed the non-toxic components that were not i included in the MSDS. The crude oil was thought to contain sulfur by their chief , chemist. High sulfur crude would be inexpensive and an obvious choice for such a j resilient seal that does not need a long life, but this sulfur and the chlorine from the PVC complicates i's destruction and shortens the functional life of the CFE and the
~ TRJ.RPT-2-03 Rev.0 l
SEG/TRJ/R-006, Trojan Second Integrated Test Data Report h@ ame aw m m I SSR stainless steel components. Since the steam reforming process gasifies any chlorine present, no chlorine is expected to remain in the process can. The experimental evidence supports the high sulfur crude additive and the zine release agent or calcium carbonate filter as illustrated in the following bullets: e The white coated lumps on the surfaces of the steam-reformed clumps are zine salt and calcium carbonate in PVC flanges e The zinc salt deposits came from the zine release agent in PVC flanges e The calcium carbonate deposits came from the calcium carbonate in PVC flanges e The evolution of extra hydrogen is associated with the white coated lumps e The evolution of hydrogen sulfide (H2S) during H 2analysis is associated with the white coated lumps e The white coated lumps are fully destroyed in the 1100*C quartz tube H2 testing
- The crude oil is the source of H 2S that is released in destroying white coated lumps during H2 analysis
- High sulfur crude contains high boiling polyaromatic sulfur compounds e The TGA/GC analysis done at IT, Corp. labs has revealed the minimum temperature at which these sulfur-aromatics can be destroyed.
e Smaller sized PVC flange material (<1/4") will steam-refonn more completely and will result in lower residual hydrogen. e The PVC flange material when cut up in small pieces and limited to 53.5 g per process can, can be processed at 100-150*F higher CFE temperatures than was used for non-PVC runs.
- Samples of the raw PVC flange material and the steam-reformed residue biased for white coated lumps were run on a Thermogravimetric-GC analyzer to obtain the gas evolution compositions at temperatures ranging from 600 to 1100*C in 50*C increments. Both PVC samples continued to outgas in small amounts above 600 C (l100 F).
e TGA/GC analysis was done on an actual radioactive sample of the fuel pool waste, revealing the same " fingerprint" of the surrogate waste suggesting very similar composition materials. The test results are best summarized by concluding that surrogate waste samples with no PVC and no Raschig rings with seven filters placed into the process can have passed the 0.05 g-moles hydrogen criterion more easily than if PVC is present. Run
#8 wnh Berl saddles easily met the 0.05 g-mole criterion by processing for 10 hours, while Run #5 without Raschig rings or saddles required processing for 27.5 hours.
l Both were biased samples. Run #1 uniformly sampled surrogate waste nearly passed l by processing for only 6.9 hours and at 1050*F CFE and steam temperature and with sample boat blank correction applied. Run #2 was not acceptable because the low flows caused by early melting increased processing time to 34 hours. The fused dross, mai reliable sample taking impossible. ! TRJ RPT-2-03 Rev.0 p
SEG/TRJ/R-006, Trojan Second Integrated Test Data Report
===@@ umo- =@.c l Next, a comparison should be made between Run #1 and Run #3, that differed j principally with PVC not included in Run #1 but included in Run #3. All other l process conditions were nearly the same. The PVC inclusion yielded a final hydrogen l
value of 0.14 g-moles H i. Longer processing time was used in Run #7, where the final steam temperature reached 1050*F and the process time of 37 hours showed fm' al hydrogen sensor readings down below 40 ppm for the fm' al 4 hours. The longer processing time did not reduce the residual hydrogen level. Further insight was obtaircd by resampling Run #3 containing PVC that had 0.14 g-moles H2and biasing the se <ple with the same fraction of white coated lumps that l was used to bias the Run #7 sample. The "Run #3 Resample" showed 1.48 g-moles H 2- Proving that the white coated lumps contained unprocessed PVC and crude oil. t Recalculating the hydrogen content by eliminating the dross weight from the i calculations significantly changed the results. Run #9 changed from 0.082 g moles to ) 0.027 g moles H i. Run #9 meets the criteria of 0.250 g moles He when five process cans am placed in a process can capsule. The method of correcting hydrogen content , l for a run is shown in Section 3.24. Analysis by IT, Corp., and the results of Run #9 showed that as particle size is reduced, the PVC is more thoroughly processed. For a given particle size, however, ! the resultant He content in the residue is proportional to the amount of PVC placed in l the process can. I l l l' I I TRJ-RPT-2-03 Rey. 0 p I ,
i SEG/TRJ/R-006, Trojan Second Integrated Test Data Report L j@@ N ICOLOGY OACW. eC, 5.0 ACRONYMS Ar Argon "7 ! Cs activated radioisotope of Cesium CH4 Methane :
- CFE Can Feed Evaporator CO Carbon Monexide i CO2 Carbon Dioxide .
EDM electron-discharge machining - g gram (s) i g-mole gram mole i GC gas chromatograph
]
H Mono-atomic hydrogen H2 Hydrogen in its normal diatomic state HO 2 water . HS 2 Hydrogen Sulfide l 1 l in w.c. inches, water column j IT, Corp International Technologies Corporation kg kilogram (s) min minute mi milliliter (s) , N2 Nitrogen l NASO,2 Sodium Sulfate I NRC Nuclear Regulatory Commission i l
- TRJ-RPT 2-03 Rev.O l
l l SEG/fRJ/R-006, Trojan Second Integrated Test Data Report - @@@ l SOENTRC ECOLOGY Geur, aC O2 Oxygen i PGE Portland General Electric, Inc. j ppm part per million POP Proof-of-Principle psi ' pounds per square inch psig pounds per square inch gage PVC polyvinyl chloride, black filter flange material
)
scfm standard cubic feet per minute . SEG Scientific Ecology Group,Inc. SSR superheated steam reformer l TGA/GC Thermogravimetric analysis by gas chromatography vol% volume percent WNP-1 WPPSS Nuclear Plant #1 l WPPSS Washington Public Power Supply System l l l 4 l TRJ-RPT 2-03 Rn. 0 Page 83 of 87 i
SEG/rRJ/R-006, Trojan Second Integrated Test Data Report cj@@@ x c ==, m .c
6.0 REFERENCES
6.1 Nuclear Regulatory Commission, Federal Code 10 CFR 71 and IE 84-72 NRC, govems the certification of conformance of the individual transportation casks requiring <5 vol % H 2-6.2 T. R. Galloway, D. F. Gagel, & N. W. Dunaway, Recent Experience With SEG Steam-Reforming of Various Types ofRadwaste, invited paper for Winter AIChE Meeting in San Francisco, CA Nov. 10-15,1996. 6.3 SEG Report Trojan ProofofPrinciple Test Report, SEGfrRJ/R-002, Rev 0, Nov. 22, 1996. 6.4 SEG Procedure Trojan Integrated Test Sampling and Analysis Plan, SEGITRJIPRO-032, Rev.1, Jun 30,1997. ; i 6.5 SEG Procedum Trojan Second Integrated Test Plan and Procedure, SEGffRJIPRO- l 033, Rev.1, Jun. 30,1997. 6.6 SEG Procedure Residue Hydrogen Analysis Protocol, SEG/TRJ/ PRO-006, Rev. O, Oct 10,1996. 6.7 IAHODA, Edward, " Reactions of Uranium Dioxide in Synthetica Process," Westinghouse Science and Technology Center, Internal Memorandum March 12, 1 1996 l l f i 1 TRJ-RPT 2 03 Rev.O Page 84 of 87 L
. SEG/rRJ/R-006, Trojan Second Integrated Test Data Report M
== me. =
l APPENDIX A - DESCRIPTION OF TEST EQUIPMENT ' l l 1 A.1 Overall Process l Steam-reforming is a two-step thermal treatment process. Hydrocarbons in the waste l are evaporated in the first-step "Can Feed Evaporator" (CFE) process unit. Within the l CFE, the hydrocarbon vapors am first exposed to superheated steam which begins the l steam-reforming of hydrocarbons into CO, CO2 , H2 0, H 2, and a small amount of CH4. The gases generated in the CFE pass through a high temperature filter and cesium trap i ! to remove radioactive particulate prior to entering the second step of treatment. In the ! I second step, the gases are heated further within the SSR, driving the steam-reforming ! reactions to completion. The reformed gases, mixed with steam, become the " syn- . gas" which is recirculated in the SSR to continue the reforming reactions to j completion. Excess syn-gas is vented through the vent to the HEPA filtemd exhaust ! system. l A.2 Component Description j 1 Drawing TRJ-S-3-03, Trojan Steam Reforming System P&lD, provides a basic schematic of the important operational features of the steam-reforming system. The heart of the system is the SSR. It is connected to a CFE with a cesium trap and filter. J j A small electrically powered steam boiler with a superheater and a nitrogen preheater i is also included. Each of these components is briefly described below. l ! A.2.1 Process Can. The process can is the actual process can planned for the Trojan Spent Fuel Pool Debris Removal Project. The can is 29 inches in height,7.5 inches in diameter, has a sealing surface machined into the bottom outside surface, and has 5 micron filters covering the top and bottom flow ports. The process can's upper filter assembly is designed with removable bolts to aid in sampling the residue for remaining hydrogen content. A.2.2 Can Feed Evaporator. The CFE consists of an approximately 10 inch diameter vertically mounted tube which is mounted inside of a shielded enclosure. A process can loaded with surrogate waste is lowered into the tube using a shielded Transfer Bell. The lower seal of the process can I mates with the seal located near the bottom of the CFE tube. This seal 4 forms the boundary between the steam inlet to the process can and the - steam chamber surrounding the process can. Weights are provided to rest i L. on top of the process can to counter the effects of steam pressure under the l process can during processing. The CFE tube and the piping supplying
- steam to the CFE are clad with pipe heaters to minimize heat losses.
j During processing, superheated steam at 1000 to 1200*F is ported into the
- bottom of the process can at the rate of I to 5 scfm. The steam and organic vapors pass through the can and exit at the top via the exit L
T1U.RPT 2-03 Page 85 of 87
. , -- . _, . - - ~ _. . - -.
SEG/TRJ/R-006, Tr:Jan Sec nd Integrated Test Dr.ta R: port M chamber. The vapors are pulled into the detoxification reactor in the SSR for conversion into syn-gas. An additional 7 to 15 scfm of superheated steam is ported to the steam chamber where it heats the process can externally and then passes through an annulus restrictor into the exit chamber. Pressure in the exit chamber and the steam chamber are maintained slightly negative. Flowrate, temperature and pressure instruments are shown in Drawing TRJ-S-3-03, Trojan Steam Reforming System P&ID. A.2.3 Cesium Trap and Filter. The Cesium Trap treats the hot vapors exiting the evaporator to minimize particulate contamination and cesium ("7Cs) ) gas carryover into the SSR. This high-temperature unit contains a sinter bonded,1-2 micron size Inconel fiber filter and a borosilicate bead cesium i trap. The cesium trap and filter medium is contained in a tube shaped l container nearly identical to the process can to facilitate handling with the i same Transfer Bell and final storage of spent filter media in similar casks as used for the process cans containing residue. The Cesium Trap is located within the same shielded enclosure as the CFE. A.2.4 Transfer Bell. The transfer bell provides a shielded enclosure with an air ) lock gate and an enclosed winch and process can grapple. The bell is used ! to transfer process cans into and out of the CFE under a nitrogen inerted atmosphere. It will also transfer process cans into and out of the glove box while maintaining the inert atmosphere. j A.2.5 Steam Reformer. The SEG Steam Reformer (SSR) contains a steam injection point, detoxification reactor, heat exchangers, blowers, and vent. i Vapors exiting the CFE and cesium trap are preheated in the reactor base heat exchanger prior to entering the detoxification reactor. The reactor is electrically heated to approximately 1600*F to complete the conversion of organic vapors into syn-gas. The syn-gas consists primarily of CO, CO2 . H 20, H ,2 and a small amount of CH,. The multi-stage heat exchangers cool the syn-gas to 300*F prior to entering the blower while preheating the vapors returning to the reactor to 1200*F. Air in leakage is controlled to maintain the syn-gas oxygen concentration below 3 percent *3% to prevent oxidation reactions. The blower provides the motive force to recirculate the gases. Excess gases not required for recirculation are vented to atmosphere via the HEPA filtered exhaust system. More detailed information is available in SEG/TRJ/ED-001, Equipment Descriptionfor the Steam Reformer. A.2.7 Computer Control System. Signals from vent gas analyzers, limit switches, thermocouples, flow sensors, pressure sensors, and current sensors are transmitted to a programmable logic controller. All operating temperatures, pressures, and flow rates are displayed for the operator's continuous monitoring at the Operating / Interface computer. In addition, Page 86 of 87
. 1 SEG/TRJ/R-006, Trojan Second Integrated Test Dato R: port i j@@@
SQDnWC ECOLOGY GMOLP. DC. the discharge gas composition is monitored continuously for hydrogen, carbon monoxide, benzene, and oxygen. The steam reformer operating l technician controls steam boiler operation, injection rates of steam and air, reactor heaters, blower speed, and valve throttle positions to maintain 1 to 5 scfm of steem flow into the process can, operating pressure in the CFE is below one atmosphere or at a negative pressure, and optimum temperatures to minimize emissions. A.3 Process Operation I Batch operations are conducted using the CFE. A process can, filled with wet i surrogate waste, is lowered into the CFE and the shielded gate is shut. The system is brought up to temperature with superheated steam from the boiler which is ported into the steam chamber surrounding the process can. When temperature rises to abovs 400*F in the steam chamber, which indicates that all standing water in the process can has vaporized, superheated steam is ported into the bottom entrance of the process can to commence vaporizing organics in the waste. Electric heaters in the detoxification reactor raise the gases to 1600*F to drive the reforming reactions to i completion. Plots of hydrogen and carbon monoxide concentrations of the syn-gas will show concentrations rising to a peak and then falling to near background levels indicating the process is completed. When all organic material has been converted, the system is placed into a cooling cycle with a nitrogen purge to lower the process 1 can temperature to enable safe handling during removal. The process can is removed under a nitrogen blanket to prevent moisture from adsorbing onto the residue. The process can is lowered into the inerted glove box where the residue is transferred into l a designated container for laboratory analysis. The steam reforming system is prepared for the next test run. l l 1 Page 87 of 87
- ...p ,
y 1 l I l 1 1 ENCLOSURE 3 l SPENT FUEL POOL DEBRIS PROJECT SEGREGATION OF MATERIAL PROCEDURE I}}