ML19221A311
| ML19221A311 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 04/30/1979 |
| From: | NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
| To: | |
| References | |
| NUDOCS 7905210248 | |
| Download: ML19221A311 (9) | |
Text
'
/~prit N
A Avio s.
11.
COLLECTION OF GASES FROM WATER
.o 1.0 Introduction Dissolved gases are obtained from water samples by pumping a vacuum over the sa=ple and stripping the water vapor from the gas stream in a cold trap. The apparatus used to separate the gases from the water and to measure the gas volume is shown in Figure 1.
The maximum volume of gas that can be collected is limited by the mercury depth in the base of the Toepler pump. This maxin.um volume is approximately 100 standard cc; however, optmum gas collection is 25 s tandard cc or less.
Avoid over-pressuring the system liecause the glass apparatus may shatter.
2.0 Apparatus The components as described in Figure 4 include a 500-ce ballant volume, a cold trap, a Toepler pump, and a gas burette with 1cveling bulb. Mercury is the working fluid in both the gas burette and the Toepler pump.
One mechanical pump is used to evacuate the gas system and to operate the Toepler pump; another pump is used to evacuate the sampling bulb.
A Pirani tube is sealed into the apparatus to determine the quality of the vacuum and to test for leaks in the system.
The small sampling bulb (approximately 5 to 10-cc) is used to separate a portion of the gas collected from the sample for analysis by mass spectrometry.
Water samples to be degassed usually are supplied in stainless steel bombs of about 500-cc, and these bombs are connected to the ballant volume by a length of flexible tubing.
Other size samples can be accommodated but require that the quantity of gas obtained be compatible with the volumetric limitations of the apparatus.
2.1 Normal Operation The sample bomb is connected to the empty ballant volume with flexible tubing. The mercury in the Toepler pump is lowered by carefully opening and adjusting stopcock D, which is cormected to the mechanical pump.
Stopcock B is then turned to connect the Toepler pump and cold trap, and the system is evacuated by opening stopcock A to the mechanical pump.
159 105 Ci 790521019 l
f
While the gas system with the cold trap is degassing, the gas burette, sampling bulb, and small section of tubing between stopcocks B and C are evacuated with another mechanical pump.
The cercury is outgassed by raising and lowering the leveling bulb while pu= ping.
Stopcock C is turned so that the gas burrette and tubing between B and C can be filled with mercury by raising the leveling bulb. The mechanical pump is lef t pumping on the sampling bulb.
When the Pirani gauge indicates that the pressure in the gas system is less than 50 u, stopcock A is closed to check the system for leaks.
Leaks are evidenced by a rapid rise in pressure en the Pirani gauge.
Stopcock D should be closed during the leak test because mercury can be forced out of the Toepler pump system into the mechanical pump if a large leak should appear in the gas system. The most common source of leakage is the connec-tion to the s.,nple bocb; the valve on the sample bomb may not be tightly closed; or water might still be evaporating from the bomb fitting under the partial vacuum.
When all leaks have been eliminated, the system is pumped down to less than 50 u, and liquid nitrogen is added to the cold trap.
(Note:
if carbon dioxide is a component of interest in the sample, dry ice and acetone must be substituted as a coolant). The pressure shown by the Pirani gauge should now be less than 10 u.
Stopcock A is closed next; no significant rise in system pressure should occur. At this stage, the mercury in the Toepler pump can be in either the raised or lowered position.
The valve on the sample boch is opened to allow water to run into the ballast volumu.
As soon as the ballast volume is partially filled, the sample boch is lowered so that no more water flows into the ballast volume (do not allow the flexible tubing to become kinked).
The ballast volume is filled only partially so as to leave considerable water surface area exposed to the vacuum system; this assures maximum release of the gases from the sample.
A rise in pressure usually is observed as the gases are released and generally is followed on the "7 eak Test" scale of the Pirani gauge since the pressures are normally higher than can be read on the two pressure scales. When no further increase in pressure is observed, the release of dissolved gases is censidered complete.
If theiteit'c2'y'har beent in the raised position in the Toepler pump, the mercury lAcWidulcmered-by carefully opening stopcock D to the vacuum pump.
Stopedck 3^1's examined to make sure it is open to the cold trap.
Approximately 1 min is allowed for the gases to come to equilibrium in the system. The gas is transferred from the Toepler pump to the gas burette by opening stopcock B to the gas burette and by raising the mercury in the Toepler pump by carefully opening stopcock D to the atmosphere. When the gas in the Toepler pump has emptied into the burette, stopcock C is turned to close off the gas in the burette.
The leveling bulb is used to equalize the pressure in the gas burette with atmospheric pressure, and the gas volume is read.
1S9 106
a A second Toepler pump strike is taken by dropping the mercury in the Toepler pump af ter opening stopcock B to the cold trap and re peating the operation as described.
Toepl er pump strokes are repeated until no increase in gas volume is observed.
As many as 10 Toepler pump stroked may be required.
The measured volum of gas is corrected c
to standard conditions with the following equation:
Pl Tstp Vstp "
VI (1)
Pstp T1 where Vstp " volume of gas at standard conditions, cc Tstp " 273.16 *K Pstp " 760 mm Hg P1 " barometric pressure mm Hg T1 ~ room temperature *K V1 " volume as gas measured, cc.
Another method for calculating the volume of gas recovered arised from the fact that each Teopler pump s t rcke in a fixed percent of the total volume of the system.
If the first Teopler pump stroke delivers 40% of the total gas in the syste then the second will dellver 40% of the residual 60% or 24% of total.
Using this re-lationship, e value for the total gas in tne sytem can be calculated from each Toepler pump strokes although it is better to take several pump strokes to minimize any forcible fractoration of the gas during release from the water or from the cold trap.
The percent of each Teopter pump stroke to the total volume must be determined for each size sample container but can be calculated from the nonmal data taken from the accumulated gas volume and pumbers of Toepler pump strokes as follows:
Toeoler Pumo St roke Gas Accumulated (cc)
Gas each Stroke (cc) 1 5.7 5.7 2
9.4 3.7 3
11.*5 2.1 4
12.8 1.3 The percent of gas remaining in system af ter each pump stroke is:
3.7 -
64.8%
1.3 - 62%
5.7 2.1 2.1 -
57% average - 60%
3.7 159 107
r The r e f o re, a pump stroke in this example is 40% of the total
.olume.
Table 11 shows a typical gas volume calculation using this relationship.
This calculated total volume also must be corrected to standard conditions.
Af ter the gas has been collected and measured, a portion is expanded into the sampling bulb for analysis by mass spectrometry.
TABLE 13 GAS VOLUME CALCULATION Toepler Pump Gas Accumulated Total-Gas Calculated Total Vol.
Strokes (ce)
(%)
(ce) 1 5.7 40 14.3 2
9.4 64 14.7 3
11.5 78 14.7 4
12.8 87 14.7 5
13.5 92 14.7 3.0 Tvoe of Samotes s
3.1 Reactor Primarv Coolant Vater Sam 61es The gases normally observed in these samples are H, He, N 2
02, Ar, and CO.
If carbon dioxide is requested in the analysis, $fy 2
ice and acetone must be used as the coolant in the cold trap.to remove water vapor.
This type of sample usually is received in a 500-cc stainless stell Hoke bomb.
A 50-m1 gas burette is used to collect the released gases.
Results are reported as standard cc/ liter of water for total gas recovered, and from the mass analysis of the gas, stan-dard cc/ liter for each of the observed components.
The volume of water for each sample is measured after recovering it from the gas collection apparatus and the sample bomb.
3.2 Miscellaneous When samples are received with no estimation of the quantity of gases per sample, the 50-m1 gas burette is used during gas collection because this burette is sultahle for volumes from approximately 1 to 50 ml.
A 5-m1 gas burette is available for use with samples known to con-tain less gas than can be accurately measured with the 50-m1 burette.
When opening any sample container to the gas collection appar-atus, care must be tak~en because a sample containing a large volume of gas under pressure coula break the glass system.
Watching the Piran!
gauge while. opening the sample container is good practice;
- although, some acceptable samples will have sufficient gas to be off scale on the gauge.
An abnormally large volume of gas will cause pressure suf-ficiant to force gas and mercury through the tube in the Toepler pump into its reservoir. Approximately 80-mm Hg pressure in the system is required or about 50 standard cc of gas if the Toepler pump reservoir is evacuated as it is during normal operation.
I f this occurs, the valve on the sample container should be closed immediately.
159 108
~
GL e
e
'h ' L 9
ve Gemau% Q N
em av s (MFMA )~D I'
/
./
J' immR
< w_ l 'u 100
- Rb h
3 L3oo m8 y,,,
a_
-3deri "7(
if Mk C
NW75 1
Q s. E c leo d )
+pa QFuchd tschelsJ) p=
- e-8 e
O' Y%
k4 %
e
<< 2 m \\
god h
- f%1 159 109
P k
m
.1
(/.--_
..%'N
-d
,5
[d(3
.I 9
2
-Q e
I
@}f
" E i
A
-a u o 3e~,p c
c.
-~
1
. ie v'. a O CD 32$
n t,-
w
-q] -
V co al o e
g
.c sim m
oE E
- S o rx c
5 N
a
~
o Y,-~.7 O o M
O u y.-
g>
'03 I
((kr@.
t O
,t--
"' N 8
pg K
N,,
(.' -
- .j j_
,i;:I' 7.- _ __-.___ - _ _ _ _ s.
g
\\
y+
in
's i
N)a t
s r
h.
q i, d.
[
D Ifi l;i l;
l.
i r.
5
~
VQ C D t _ -. D. I, 1 - - _. - - - -- _
)
p t; no 1
\\
r,-
5 d. d Pg an 53 D
@w] -.
.b o J
gF
!j c
$y..
3 g_ E g
i
,j u
'0
- d 5 iY, C
~ ;:~
h*
._J ll J
O. c u.
m ~
~
- r
!jij t
q
>s..
D 3
a-a s
i m d t) n I
hv, w.2 ~.
.aO
,,s up
.u.
- a. ps.
- r.,
p l
1
-q a
N T
3 lI
.,,~...
\\ \\
~ a n,
ss
- I c) w
- a.,
w i
. CL, -(
i g
J '- < Gi I
n;
[/
mg 4'
>~
O O 'm 3).
nl g
hs I
0 w
@cf h j.&
'57'D ;... l...,..........,, i, Im i i i wm _.e-~.g' \\
.El i
cn u_
s
(
4 c
t 7
~.
u 4:3
=n
.q\\
x gg
- o. -
5 r=5 ED
~u m
CD t' c O
/
Mit
<b
~ ~ ~ ~f_. N ffi (O
/
ls!
\\., I,./
[
p4 i
c) g<
_.f
/,
= -Q h v +I
~~2 C
s E
O ~3
$(g; o
o t';c gs ww w
L'd J
4 flA o
>=J 159 110 na
.sa
The sample arrived in Idaho Falls at 0640 hrs.
MST on 4/14/79 vie air t r an s por t a tion.
Figure 1 illustrates how the sample was packed, with activity readings.
Srcears of the air craf t and stripping drum ind Lcated no external contamination.
At 0715 the sample Icf t the Idaho Fall Airport, it arrived at INEL (CPP-602) at 0345 hrs.
Upon arrival at CPP the sample (shipping drum) was off-loaded with an overhead hoist onto a four-wheeled dolly and transferred to lab 103C. At no time was the shipping drum allowed in a nonvertical position.
The ikb 103C, the shipping drum was opened and the 8" diameter pipe lifted into a chemical fume hood.
In the fume hood a sample of the organ atmosphere within the 8" pipe was taken.
The method used for taking this sample involved when value 8 was opened on the TMI sample (
1105 hrs MET) caly a small rise in pressure was observed on the Pirani gauge (Figure 2) which indicated a small amouct of gas compared to the ATR sample.
Consequently, the TMI sample gas was c.ollected for approximately 1 hr er 10 Teopler pump strokes.
The volume of gas collected was 0.8 +.2 cc and had the fo11cving composition:
H2 1.5 +.1*
He 0.01 89.4 +.1 N2 02 8.1 +.1 Ar 11.00 + 0.02
- Uncertainties are based on more spectrometry measurements only.
The more spectrometry analysis is detailed in reference ICP-1031.
A ga==a-ray pulse height analysis of the TMI gas sample indicated no observable activities above background.
159 111
APPENDIX A Volume Measurement Due to the sensitive nature of the analysis, we initially attempted to be extremely conservative and include all possible sources of error (no matter how remote Ctey may be).
The measurements indicatcd a total gas volume of 0.8 cc, but to be conservative a range of 0.8-1.7 cc was reported.
The initial measurement of the total gas volume performed at the gas collector is eupported by the pressure measurement performed when the has from the 10 cc simple bulb was expanded into the mass spectrometer inlet.
Based upon detailed analysis of the data, we conclude that the best estimate of the total gas in the 31.5 mr bomb is 0.8 std. cc and that this value is good to + 30.
Air Inleakage We estimate that the air inleakage to the TMI sample was less than 0.07 std.
cc.
This estimate is based on the results using the ATR sample.
To obtain an upper ibnit for air inleakage, it was assumed that all the oxygen detected in the ATR sample was due to air inleakage.
Pumping time for the ATR sample was about 15 minutes.
s 159 112
t APPENDIX B Leakaee to Cover Gas Based on evaluation of the cover gas analysie, the maximum detectable hydrogen Icakage from the 31.5 m1 bomb to the cover gas is estimated to be 0.2 std. ec.
This is based on the detection limit for hydrogen (i.e., 0.0047.) and the assumption that the gas volume in the 8 inch pipe was 5000 cc.
Max. H2 0.00004 (150)
= 0.2 cc 150/5000 159 113
___ _ _ _ _