ML20058E227
| ML20058E227 | |
| Person / Time | |
|---|---|
| Site: | Framatome ANP Richland |
| Issue date: | 11/30/1993 |
| From: | Kohler J, Mclees R, Washington L SIEMENS CORP. |
| To: | |
| Shared Package | |
| ML20058E224 | List: |
| References | |
| EMF-1505, EMF-1505-R02, EMF-1505-R2, NUDOCS 9312060218 | |
| Download: ML20058E227 (21) | |
Text
-
s.
SIEMENS EMF-1505 Revision 2 ha Siemens Power Corporation Occurrence of Free Moisture in The 63-Cubic Foot Day Mixer l
November 1993 i
. +.
f1 q
[
Rt.1 s Siemens Power Corporation y{!
i Nuclear Division L
9312060218 931112 E
{DR ADOCK0700g7
EMF-1505 Revision 2 1ssue Date: 11/8/93 Siemens Power Corporation Occurrence of Free Moisture in The 63-Cubic Foot Day Mixer Prepared by:
b h
R. B. McLees, Manager Process Engineering Manufacturing Engineering
} e [ n '/ ] )
l 7
J. E. Kohler Process Engineering Manufacturing Engineering i
A L. D. Washingtori /
Process Engineerimg Manufacturing Engineering November 1993 i
I l
l
EMF-1505 Revision 2 Pagei-Occurrence of Free Moisture in The 63-Cubic Foot Day Mixer CHANGES THIS REVISION Section
, Nature of Change Engineering Justification All Sections and Removed proprietary information.
Not required.
Figure 1,0-1 l
)
l i
EMF-1505 Revision 2 Page ii Occurrence of Free Moisture in The 63-Cubic Foot Day Mixer TABLE OF CONTENTS Section Pace
1.0 BACKGROUND
1 2.0 REVIEW OF POTENTIAL SOURCES OF HYDROGENOUS MATERIAL 2
2.1 In Leakage From the Gear Box and/or Swing Arm 2
2.2 Moisture Entry by Uncontro: led Access 2
2.3 Entry of Moisture with the Conveying Air 4
2.4 Entry of Moisture with Blend Feed 5
3.0 CONCLUSION
S 6
4.0 RECOMMENDATIONS 7
LIST OF TABLES Table Pace 2.4-1 Moisture Balance - Powder Lot 492-01 Based on Hydrogen Analyses 8
2.4-2 Moisture Balance - Powder Lot 492 Based on Moisture Analyses.
9 2.4-3 U-235 Measurements 10 LIST OF FIGURES Fioure i.0-i Biender System Schematic 11 2.3-1 Blender Test Matrix 12 2.3-2 Blender Ter. Data 13 2.4 1 Inlet Ports and Top of Auger / Lower Portion of Blender 14 2.4-2 VAC-U-MAX Filter Housing 15
EMF-1505 Revision 2 Page 1 Occurrence of Free Moisture in The 63-Cubic Foot Day Mixer
1.0 BACKGROUND
Siemens Power Corporation - Nuclear Division (SPC) completed installation, in June 1992, of a 63-cubic foot capacity, planetary auger, conical blender manufactured by J. H. Day Company.
This mixer was selected to homogenize various combinations of uranium oxide powders, based on over 15 years experience at SPC with similar, but smaller, machines. On completion of installation, the Startup Safety Council approved operation of the blender on a limited basis, under Engineering control, to perform qualification runs. After completion of the qualification blends and execution of punch list items, the Day mixer installation will be re-reviewed by the Startup Safety Council and it will be placed into full production as appropriate.
Following completion of the second qualification blend (=5709 kgs), evidence of free moisture was discovered in the blender when filter covers were removed for inventory cleanout. Four filter housings, which are used to separate the pneumatic conveying gas (room air) from uranium oxide powder during loading, are mounted on the top cover of the Day mixer (see Figure 1.0-1).
The presence and quantity of a hydrogenous material within the confines of the blender are significant, since criticality safety of the system is, in part, based on strict control of hydrogen (moderation control) within the unit and its e > ents. This control is achieved by limiting the qt atity of moisture in the feed to 3500 ppm (hydrogen limit 390 ppm), and by controlling paths by which hydrogenous materials can enter the 63-cubic foot mixer.
Four means by which hydrogenous material can enter the 63-cubic foot Day mixer include:
As a lubricant leak from the gear box and blender screw swing arm, Uncontrolled access to the various blender openings, With the pneumatic conveying gas, and As part of the powder feed to the blender.
The potential for each of these four routes was examined in detail, as discussed below, and each was considered as part of the safety review prior to the blender installation.
I
EMF-1505 Revision 2 Page 2 2.0 REVIEW OF POTENTIAL SOURCES OF HYDROGENOUS MATERIAL 2.1 in Leakaae From the Gear Box and/or Swina Arm Both the gear box mounted on top of the Day mixer and the swing arm inside the mixer are lubricated and can, on significant wear of component seals, leak lubricant into the blender cavity. This grease and oil can also be moved into the filter housings along with the transport air stream.
As a precaution agairist this occurrence, SPC lubricates all Day mixers with halocarbon oil and grease, compounds developed for the nuclear industry as hydrogen free lubricants with little moderating capability. SPC supplied these materials"' to J. H. Day Company for installation in the 63-cubic foot blender, after standard lubricants were removed from bearings and gear assemblies by the vendor. Installation of halocarbon lubricants was witnessed by J. H. Day Company supervision and certified for SPC.
The material removed from the filter housing area had neither the appearance nor consistency of oil or grease, and was analyzed as having significant hydrogen content.
No evidence of major lubricant leakage was found at any o'f the seal areas inside the blender.
Based on these observations and analyses, it was concluded that lubricant leakage was not the source of hydrogenous material found in the filter housings.
2.2 Moisture Entry by Uncontrolled Access Moisture or other hydrogenous material could potentially enter the 63-cubic foot blender through any of several openings into the blender. These include the filter housings and associated conveying gas exhaust lines, the powder conveying lines, the dump valve at the bottom of the blender,'he sample port located on the blender side, and the vacuum relief valve openings on the blender lid.
An evaluation was performed to determine the probability of each of these openings as an entry point for the observed moisture. The blender filter housings, associated vacuum hoses and valves, and canister filters immediately ahead of the vacuum pumps were Halocarbon oil series HC-56 and halocarbon grease series25-10M.
EMF-1505 Revision 2 Page 3 disassembled and examined for signs of water. The clear vinyl hoses and four filter canisters showed no evidence of water entry, such as discolorations or wet or dry powder caked on the walls or on the cloth filter socks. Potential sources of water above the blender and in the area of the pumps and filters have been contained or eliminated.
The vacuum relief valves were examined and found to be shut; and from evidence discussed in Section 2.4 below, these spring loaded valves, which can only open during the powder conveying cycle, were determined not to be viable entry points for water into the blender. A sample port, located on the blender wall approximately five feet above the plant floor, was also eliminated from additional consideration. It was properly capped during the loading, blending, and unloading sequence.
The four filter housings located on the blender lid are enclosed in hoods and no evidence of water was found in or around the hoods. The fi!!er covers remained in place and sealed until after the blending operation was completed, and the mixer was unloaded.
g These two openings were eliminated as the moisture entry point (s).
The powder conveying lines terminate in discharge hoods below the Line 2 and Line 3 blenders. These hoods are designed for containment of airborne uranium dust during downloading of the blenders, and also limit the potential for uncontrolled moisture entry s
into the 63-cubic foot blender. As indicated above. in consideration of the vacuum relief valves, entry of significant unectdrolled :
itities of water during the blender filling operations is not supported by examination of the emptied blender.
It was concluded from these examinations that i ;ontrolled entry of water into the 63-cubic foot blender through various openings was not probable under conditions existing during the blend qualification run.
EMF-1505 Revision 2 Page 4 2.3 Entry of Moisture with the Conveyino Air The third scenario for build up of water in the blender is by condensation of moisture from the conveying gas. In order to test the potential for this occurrence a test matrix was developed to reproduce the blender loading and unloading sequence without introduction of uranium oxide powder to the system. This test (see Figure 2.3-1) included installation of thermocouples inside and outside of the filter housings on the blender lid.
Temperature was also measured just above floor level at the bottom of the 21-cubic foot blender where powder conveying air" originates. This position was also used for measurement of dew points in the room, and a second dew point sampling point was established inside the blender at the filter housing.
The blender and associated pneumatic conveying system were dry cycled over a period of five days to simulate the load, blend, and unload sequence followed for the qualification lot. The data from these measurements is plotted as Figure 2.3-2.
In addition, the filter housing covers were removed after each cycle segment, and the housing interiors were examined for signs of moisture condensation.
None were observed during the entire test procedure.
These data indicate that the dew points of the air inside the blender and in the room outside the blender generally maintained within a 10 F band and showed no major swings with time or with system operation. The blender dew point remained consistently below that of the room; however, this can be attributed in part to the differences in the measuring equipment. A Panametrics moisture analyzer was used to sample the blender filter housing and a hair hydrograph was used to measure room conditions.
Temperatures top to bottom and inside the blender maintained well above the dew points inside and outside of the blender. From these observations it was concluded that condensation of moisture from the conveying gas stream is not probable under typical summer conditions. Winter conditions, which tend to further dehumidify the room air, make condensation in the 63-cubic foot Day blender even less probable.
The conclusion from this test was that a source existed other than conveying air for the moisture observed in the blende'r.
EMF-1505 Revision 2 Page 5 2.4 Entry of Moisture with Blend Feed The fourth path for entry of hydrogenous material into the 63-cubic foot Day mixer is mixed with the powder feed. The SPC conversion process typically produces calcined powder which ranges in moisture content from 500 to 1500 ppm"' As shown in Table 2.4-1, hydrogen analyses for the sub lots (481-01,02,03, and 04) which were blended to form the 5709 kg lot,492-01, ranged frcm 101.2 ppm to 128.9 ppm. The equivalent moisture contents calculated from these values ranged from 910 ppm to 1160 ppm. The weighted average for the blend feed was 1061 ppm, and the total equivalent moisture content calculated from this average was 6.06 kgs of water
- Table 2.4-2 summarizes the results of moisture analyses representative of the four sub lots. These value range from 758 to 923 ppm, and the moisture content of the blend, based on these a.ialyses was 4.90 kgs of water
- These data compare well and are within the accuracy range of the two analytical methods used.
A total of 5559 kgs of uranium oxide, were discharged from the blender. Moisture analyses from 5 -containers selected at random from the lot averaged 820 ppm.
Hydrogen analyses for this same material averaged 110 ppm (990 ppm moisture equivalent). In addition.150 kgs of powder were removed from the lower blender walls and auger. Moisture analysis for this material, after homogenization, was 1436 ppm, and hydrogen analyses reported a level of 207 ppm (1863 ppm moisture equivalent).
By the two analytical tr,ethods, moisture and hydrogen, the calculated water content of the blended powder lot was 4.77 and 5.77 kgs, respectively. Based on the most conservative technique, hydrogen analyses, the water lost during the blending and down loading operation was 0.29 kgs.
Uranium ovide powder produced by the ammonium diuranate (ADU) process may oxidize after it is discharged from the calciner. This oxidation step, which liberates heet, is
")
See ANF-1267(P), " Qualification Report - Phase 1", Figures 13 and 13A.
Reported accuracy of hydrogen analysis is 18%.
Reported accuracy of moisture analyses is 20%.
EMF-1505 Revision 2 Page 6 routinely observed in the hoppers and blenders, and is dependent on the surface area and activity of the powder, powder configuration, and temperature, and the availability of oxygen. The occurrence of oxidation in a powder blend is not considered deleterious to the process, since it has been shown that oxidized powder can increase the green strength of pressed pellets.
The heat from this exothermic reaction is adequate to raise powder temperatures well above 100 C, and drive off moisture mixed with the powder. Water vapor driven from UO, powder can collect o, cool metal surfaces such as the blender wall, the auger, and in the filter housing above the blender lid.
Energy introduced to the powder while blending may be trapped in the static bed by the insulating characteristics of fine particles, increasing the tendency of the powder to I
oxidize, generating more heat, and forcing moisture out of the powder bed. This scenario is compatible with visual observations and moisture measurements.
Visual observations of the blender interior indicate that moisture migrated to the blender walls, as evidenced by the compacted powder layer (see Figure 2.4-1). The elevated moisture level (1436 ppm) further suppora this postulation. Moisture levels in excess of 50,000 ppm were measured in the small quantity of uranium oxide residue found in the filter housings. This area also had the visual appearance of having been wet (see Figure 2.4-2).
Measurement of the U-235 content of the blended powder, discharged freely from the mixer, and of-the compacted layer on the blender walls is presented in Table 2.4-3. The close agreement in these analyses provides strong evidence that the moisture migration occurred after the blending step, in the course of the blender downloading operation.
During loading of the blender, a high volume of conveying gas is moved through the blender, and this flow, coupled with the cyclic evacuation"' of the blender, would sweep away any moisture liberated from the powder bed.
The blender is cycled between atmosphere pressure and minus 13 inches Hg at 1% to 2 minute intervals.
EMF 1505 Revision 2 Page 7
3.0 CONCLUSION
S 3.1 Analyses and visual examinations of the blender and its contents indicate that moisture observed in the filter housings and in the powder layered on the blender wall and auger originated in the incoming uranium oxide powder.
3.2
'A hydrogen material balance performed on the powder shows that approximately 0.29 kgs of water was removed from the powder during transport and blending.
3.3 Isotopic uniformity indicates that the moisture was liberated from the warm powder bed after completion of the blending step or during the final stages of blending.
3.4 The most probable source of the heat required to cause moisture migration is through oxidation of the UO,. This phenomenon is common and poses no threat either to the equipment or the quality of the product.
4.0 RECOMMENDATIONS 4.1 Equip the blender with a dry nitrogen sweep to remove moisture during the blending and blender download step and control dew point within the blender.
i j
4.2 Cycle the blender auger at intervals during downloading to disperse any warm powder in the blend center.
)
)
EMF-1.505 Revision 2 Page 8 TABLE 2.4-1 MOISTURE BALANCE - POWDER LOT 492 BASED ON HYDROGEN ANALYSES SUB LOT NO.
HYDROGEN EQUIVALENT WEIGHT"'
WElGHTED MOISTURE (ppm)
MOISTURE (kgs)
MOISTURE CONTENT (ppm)
CONTRIBUTION (kgs)
(ppm) 481-01 110.6 995 1579 275 481-02 125.4 1129 1548 306 481-03 128.9 1160 1557 316 481-04 101.2 910 1025 163 TOTAL 5709 1061 6.06 FINAL LOT NO.
HYDROGEN EQUIVALENT WEIGHT WEIGHTED MOISTURE (ppm)
MOISTURE (kgs)
MOISTURE CONTENT (ppm)
CONTRIBUTION (kgs)
(ppm) 492-01 CONTAINERS 109.7 987 555.
961 RESIDUE 207 1863 150 49 TOTAL 5709 1010 5.77 WATER A = 0.29 kg 14 kg MUF prorated.
s EMF-1505
~
Revision 2 Page 9 TABLE 2.4 2 MOISTURE BALANCE - POWDER LOT 492,
BASED ON MOISTURE ANALYSES STARTING LOT MEASURED WEIGHT" WElGHTED MOISTURE MOISTURE N O.
MOISTURE (kgs)
CONTRIBUTION CONTENT (ppm)
(ppm)
(kgs) 481-01 923 1579 255 481-02 891 1549 242 481-03 758 1557 207 481-04 859 1025 154 TOTAL 5709 858 4.90 l
FINAL LOT NO.
MEASURED WEIGHT WEIGHTED MOISTURE MOISTURE MOISTURE (kgs)
CONTRIBUTION CONTENT i
(ppm)
(ppm)
(kgs) 492-01 l
CONTAINERS 820 5559 798 RESIDUE 1436 150 38 TOTAL 5709 836 4.77
-a WATER A = 0.12 kgs" l
l n
14 kg MUF prorated.
Apparent discrepancy due to round off.
EMF-1505 Revision 2 Page 10 TABLE 2.4-3 U-235 MEASUREMENTS SAMPLE SOURCE U 235 CONTENT Sub Lot 481-01 3.058 Sub Lot 481-02 3.596 Sub Lot 481-03 3.595 Sub Lot 481-04 2.706 to 3.601 Blender Wall #t1 3.412 i
Blender Wall #2 3.416 l
Auger #1 3.410 Auger #2 3.411 Blender Residue (150 kgs) 3.407 i
Blended Lot (5559 kgs) 3.409
O To Sock Filters and Vacuum Pump
/
n, o
C1 g To Sock Filters and Powder Feed 0 Vacuum Pumps Q
0 0
9 9,
Q 0 0 o o.
o o o O o
o O
O o
o
's
{'
0~
)
/'
r' O'
""S*
0e f
)
Processes n_.
V
.h 1
D y
1 1
1.
Hopper 6.
Locked Valve 2.
Sample Port 7.
Powder Ccnveying Line 3.
21-Cubic Foot Day Mixer 8.
Powder Feeder 4.
Sample Port 9.
63-Cubic Foot Day Mixer 5.
Filter Housing 10.
Vacuum Flclief Valve 289
'8 i 9 5"
m"af Figure 1.0-1 BLENDER SYSTEM SCHEMATIC
EMPTY MEGA-BLENDER TEST MoniterDewPointandTemperaturesatExhaustFilterBogs.
llse attach 5d dato sheet to record the den point and temperature (s) at the N.W. filter Housing.
(Dew point and temperature inside housing plus tmperature in the com outside blender.)
w DAY l DAY 2 DAY 3 DAY 4,5, & 6
=
^
r 3,
3 I
6m 5 pm 9m 11 pm 2m Run discharge to to to to Voccuta from 9m 8 pn 1 pn 2 on Run blender for Blender 53 hours6.134259e-4 days <br />0.0147 hours <br />8.763227e-5 weeks <br />2.01665e-5 months <br />.
Vaccuta Vaccuta Vaccutn Vaccutn 45 min.
Transfer Transfer Transfer Transfer a
frm frm frm frm MegoBlender Line 2/3 Line 2/3 Line 2/3 Line2/3 to Hauto Mixer Nauto Mixer Nouto Mixer Nauto Mixer Line2/3 prep.
to to to to 3
Mego Blender Mego Blender MegoBlender MegaBlender h
a a
A V
Y Y
Y Y
Y l
Open filter llousing(s) and inspect for moisture collection.
(Take photographs if moisture is seen.)
Test to begin Tuesday. 7/28/92
- Mop the floor at the inlet vaccuta feeder point to the blender during the last hour of the test.
es.g y;p m
NOTE: Clive francis hos checked instrments before test stcrts. Relative htnidity and temperature recorder are installed 15y at the in-feed volve hood.
"g y
Al Ficure 2.31 1
<,6-l
EMPTY MEGA BLENDER TEST DATA 100 r
1 3
l T' L o.J 3" Load 4" Lo.J 01.aJ
, '* L. J k
- o... uae A
4 s
/
l
./
%d. f 80 e
IN g 'V [
l
/
7
\\
+
l l
+f I
70 4r A4
\\,,._g a
3 l
~
/
'tmw aA n
W l
40 07:48 l 06:15 l 08:15 l 17:30 l 19:30 l09:00 l11:00 l 13:00 23:00 l 01:00 l 03:20 l 14:00 l 23:30 l 09:30 f 19:30 l 05:30 l 15:30 i
i s
, 02:43, 07:15 09:15 18:30 20:30, 10.00 12:00
{ 00 02 02:00 11:00 18:30, 04;30 14:30 00 30 10:30 7/3 7pg a 7/J e rg a Sample Time ren eg a visorsa 7' 8"* a esse *a j
O dew PT. til OLENDER I
+ tt451DE TEMP.
o IOP OUTSIDE TEMP.
a DEW Pl. OUTf DLENDER X
DOTTOM OUT TEMP.
oxm E$g mo 7'
Figure 2.3-2
- 5~ G wa o 10 W (S 6
- EMF-1505 IRevision 2 "Page 14
\\
Y l'
h e
a INLET PORTS AND TOP OF AUGER l
l i
t "...,*#'g - - m,.,k',,.'sg
_,1-7
<s-w
_,.,7.,
.'8 j_f, ;g'h.
- N 34
,{..,
' $ W', ; '%'s ' ? * ~17gg' i.p W* '
u
/;o
, 7,
..,a;< f ; 9.
~ ~ ^ '
~ '
- c e 1
~
> :m~
l
, fQQD -
2.1,gign g, es,
%t
_, [ _ _ _, _ _, 3 -
LOWER PORTION OF BLENDER FIGURE 2.4-1
EMF-1505 Revision 2 l
Page 15 1
A.
/.,-
s
/
t 9
j
'!)
.': g.
)..;'t.. 'j
.')
(
'4 Qp
. g _,.-
'V
'N.
- qa
- h. '
.M c..
(;.
O ca, e'
=
g 4
h 44-
,s.,
e 4.
u.
p
.e
.n y; ;} ?[.~: y '-
P
, m. gs FIGURE 2.4-2 VAC-U-MAX FILTER HOUSING
EMF-1505 Revision 2 issue Date:
11/8/93 Occurrence of Free Moisture in The 63-Cubic Foot Day Mixer Distribution:
L J.
Maas (2)
C. D, Manning j
T. C.
Probasco I
\\
..