L. K. Heung and M. L. Rhoden
Westinghouse Savannah River Company
Aiken, SC 29808

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

This report has been reproduced directly from the best available copy.

Available for sale to the public, in paper, from:  U.S. Department of Commerce, National Technical Information Service, 5285 Port Royal Road, Springfield, VA 22161,  phone: (800) 553-6847,  fax: (703) 605-6900,  email:  orders@ntis.fedworld.gov   online ordering:  http://www.ntis.gov/support/index.html

Available electronically at  http://www.osti.gov/bridge/

Available for a processing fee to U.S. Department of Energy and its contractors, in paper, from: U.S. Department of Energy, Office of Scientific and Technical Information, P.O. Box 62, Oak Ridge, TN 37831-0062,  phone: (865 ) 576-8401,  fax: (865) 576-5728,  email:  reports@adonis.osti.gov

A tritium stripping and recovery system that supports the operation of 800 m³ of glovebox space is discussed. The system has been in successful operation since 1994. It continues to maintain the glovebox atmosphere at less than 0.1 Ci/m³, and has recovered grams of tritium from accidental releases.

The Replacement Tritium Facility (RTF) at Savannah River Site has been in continuous operation since its startup in January 1994. The Facility processes kilograms of tritium each year. All tritium handling process equipment is installed inside nitrogen filled gloveboxes. Tritium released to the glovebox atmosphere through normal and accidental operation is recovered using tritium strippers. The stripper systems have maintained the glovebox atmosphere at less than 0.1 Ci/m³ and recovered grams of tritium over the years.

The process equipment of RTF is enclosed in 30 gloveboxes, total volume of 800 m³. The gloveboxes are separated into two groups. Each group is supported by a primary stripper system (P1 and P2). Each primary system circulates the glovebox atmosphere (nitrogen) at a rate of 170 m³/hr, giving an average residence time of 2.4 hrs. A third identical stripper system called the secondary striper (SS) is reserved for accidental tritium release in any of the gloveboxes. In such an even, the particular glovebox is isolated from others and is placed in the SS system. The residence time is therefore reduced to an average of 9 minutes. The objective is to reduce the tritium activity back to the normal low level within 8 hours. The SS system is also used to handle gloveboxes that show persistently high tritium concentrations.

Each stripper system consists of a catalyst bed and three molecular sieve (zeolite) beds. Only one of the 3 zeolite beds is online at any giving time. The other two are either being regenerated or on standby. The volume of each zeolite bed is 0.16 m³ holding 104 kg of zeolite (type 3A), giving a residence time of 3 seconds. The catalyst bed is 0.14 m³ containing 130 kg of palladium on alumina catalyst giving a residence time of 1.5 seconds at 400 ^oC.

A fourth stripper called the purge stripper (PS) consists of three 0.04 m³ beds each holding 27 kg of palladium deposited on zeolite type 4A, operating at ambient temperature. One of the 3 beds is online at any giving time. The purpose of this PS is to further strip the purge nitrogen taken from the primary strippers. This nitrogen purge is necessary for the control of pressure and oxygen level in the gloveboxes. The purge rate (total for all gloveboxes) is approximatley 8.5 m³/hr depending on atmospheric pressure changes and oxygen level in the gloveboxes. The nitrogen atmosphere in the goveboxes is maintained at a negative pressure down to 124 Pa (0.5 inch water). A flow diagram of the entire glovebox-stripper system is shown in Figure 1. Other aspects of this system have been discussed in earlier publications.^1,2

The design target performance for the gloveboxes atmosphere is: tritium < 0.1 Ci/m³; oxygen < 1 vol% and moisture <200 ppm.

Figure 1. Flow Diagram of the Glovebox-Stripper System

The glovebox atmosphere is nitrogen. Tritium, moisture and oxygen are considered impurities. Tritium is from the process equipment through leaks and permeation. Moisture and oxygen are from the room ambient through leaks and permeation of the gloveboxes. Tritium is oxidized and captured together with the moisture by the zeolite beds of the stripper system. The captured, tritium containing moisture is recovered from the zeolite bed by heat and hot metal reduction. Oxygen is purged. This scheme of impurity source and control is shown in Figure 2.

Figure 2. Impurity Source and Control

Leaking from ambient into gloveboxes The amount of ambient air leaking into the gloveboxes can be calculated from the nitrogen purge rate and the oxygen concentration in the purged nitrogen. For the RTF system the purge rate is approximately 8.5 m³/hr (354 mole/hr) and the average oxygen concentration in the purged gas is 0.8 mole%. The oxygen purge rate is calculated to be 2.83 mole/hr, or 24.8K mole/yr. If the purged oxygen is solely from leaks and the leaked air has the same composition as that in the room (21% oxygen, 1.5% H₂O at 60% relative humidity), the air leak rate will be 13.5 mole/hr. The moisture leak rate will be 0.20 mole/hr, or 1771 mole/yr.

Permeation of moisture through the glovebox barrier can be calculated from the amount of hydrogen recovered from the zeolite beds. On average zeolite bed recovery generates approximately 150K STP liter/yr (6.7K mole/yr) of hydrogen. The concentration of tritium in the recovered hydrogen is approximately 0.01 mole%. The recovered tritium is therefore 0.67 mole/yr which is very small and insignificant for the purpose of hydrogen mass balance. All the hydrogen recovered is either from leaks or permeation. We have calculated above that the amount of moisture leaking into the gloveboxes is at most 1,771 mole/yr, that accounts for 26 mole% of the total hydrogen recovered. The balance of 74 mole% has to be from permeation. Permeation accounts for most of the hydrogen source. Leak tight glovebox is important to oxygen level, but is less important to moisture level. This overall material balance of the impurities in the RTF glovebox-stripper system is summarized in Table 1.

In normal operation each of the primary stripper system circulates the nitrogen at 170 m³/hr through 15 gloveboxes and strips the moisture containing tritium. The gloveboxes are connected in parallel and the flow is distributed among them according to the sizes and the anticipated tritium release rates. Ion chambers measuring the tritium concentration are located at the inlet to the catalyst bed, the outlet of the zeolite bed and the inside of the gloveboxes. Concentration at the inlet is in effect the average of the 15 goveboxes. Tritium and moisture concentration at the outlet of the zeolite bed is the lowest when the nitrogen is circulated back to the gloveboxes. Daily readings of the outlet and inlet ion chambers over a period of 3 years are plotted in Figure 3.

Figure 3. Tritium Concentration at Outlet and Inlet of Stripper

For each stripper system the data roughly consists of a series of curves that are flat at the beginning then curve up near the end. These are the periods from when a newly generated zeolite bed is placed online to when it is saturated. The shape of both inlet and outlet data look similar indicating the inlet (glovebox atmosphere) concentration responses to the outlet (return to glovebox) as expected. The outlet concentrations of P1 and P2 stay well below 0.05 Ci/m³, and that of SS is slightly higher but only get above 0.05 in a couple of occasions. This indicates that the strippers maintain their tritium stripping capability very well. On the inlet side, the concentrations of P1 and SS have increased gradually, approaching or getting above 0.1, while that of P2 remained well below 0.05. For P1 and SS, the inlet concentration has stayed at more than double that of the outlet for months. Are the ion chamber readings real? If real, the zeolite beds of these two systems would have captured and recovered much more tritium. To see if this is true, the amounts of tritium captured and recovered by the zeolite beds for all three systems during the same time span are calculated. The amount captured is calculated by multiplying the difference of inlet and outlet concentrations by the circulation rate. The amount recovered is calculated from the tritium concentration in the recovered gas during zeolite bed regeneration. Note that inlet /outlet concentrations are measured by ion chambers, and the concentration in the recovered gas is measured by mass spec. If both measurements are correct, the captured and recovered should be close to be the same. The calculated results in Table 2 show large differences for all three systems.

For both P1 and SS, the captured value is about 15 times larger than the recovered value (156 vs. 9.5 and 398 vs. 28.6 Ci/day). For P2, the captured value is slightly negative while the recovered is positive (-1.3 vs. 19.7 Ci/day). Two questions can be asked here. Which value is more real, and why P2 is different than the other two. To answer the first question, let us introduce another set of data. On December 2000, gas samples were carefully taken from three gloveboxes and their tritium concentrations were measured with a "clean" ion chamber. These ritium concentrations are compared in Table 3 with the readings of ion chambers inside the gloveboxes. The results show that the glovebox ion chambers read much higher values than the clean ion chamber did on the samples. The ratio of the values varies from approximately 50 to 800 depending on concentration (Table 3). This indicates that the glovebox ion chambers are given false high readings. It is well known that ion chambers carry memory effects. Once they are exposed to higher tritium concentrations they cannot recover to the clean state completely, particularly in dry conditions. There is little question that the inlet ion chambers of P1 and SS have been recording false high readings due to memory effect. The inlet ion chamber of P2 showed much less memory effect for reasons to be discussed below. The three data points in Table 3, though meager, fit a linear equation quite well. This equation may be used to convert the glovebox ion chamber readings to more true values. If one uses the lowest factor of 50, for a true concentration of 0.1 Ci/m³, the false reading would be 5 Ci/m³. It is therefore safe to conclude that the true values in Figure 3 are less than 0.1 Ci/m³ that is the design target limit. Longer term, ion chambers less affected by contamination should be used.

Hydrogen recovered from the zeolite beds include all three isotopes, but in terms of mass, practically all is protium from the moisture outside the gloveboxes. Based on the days online, the average tritium and hydrogen recovered from the zeolite beds of each of the 4 stripper systems over a period of approximately 1000 days are given in Table 4. It is not surprising that the SS system recovered the most tritium. It is surprising though that the P2 system recovered twice as much as the P1 system. Remember that P2 read much lower tritium concentration at the inlet that P1. The explanation is on the amount of moisture getting into the system. Table 4 shows that P2 gets 2.5 times as much moisture as P1. The higher moisture level has apparently effected the ion chamber in the P2 system to read more true values.

Gram quantity of tritium was accidentally released from the process in a particular glovebox at one time. Tritium activity in the box shot up to 2340 Ci/m³. The glovebox was immediately isolated and put on the SS stripper system as planned. The tritium activity decreased rapidly to about 1 Ci/m³ in less than 1.5 hours, and to a near steady reading of 0.1 Ci/m³ in 8 hours meeting design expectations. Getting back to the original reading of 0.05 Ci/m³ took five days. Positive aspect of this incident is that all the released tritium was captured and recovered. The tritium concentration response as a function of residence time is shown in Figure 4 in comparison with calculated results using an ideally mixed model. The ideal model shows a quicker decrease in concentration. This is not surprising since the glovebox atmosphere is not ideally mixed, and the ion chamber reading might have inflated by the memory effect. The model equation is as follows:

(C-Ce)/(Co-Ce)=exp(-F *t/V)
where C=T2 concentration in glovebox, Ci/m³, Ce=stripper outlet concentration, 0.05 Ci/m3, Co=glovebox T2 concentration at time zero, F=flow rate, 170 m³/hr, V=net glovebox volume, 13 m³, residence time=F/V=0.0765 hr=4.56 minutes.

Figure 4. Response of Glovebox Tritium Concentration
during Gram Quantity Tritium Release from the Process

The glovebox-stripper systems capture and recover tritium as expected. Their performance has maintained for more than 7 years. Tritium captured and recovered averages 60 Ci/day. Accidental release of tritium from the process is effectively confined and recovered. Ion chamber measurement of tritium concentration is complicated by the memory effect. The reading tends to increase over time but the actual level may not have changed. Moisture appears effective in reducing the memory effect in ion chambers. Ion chambers with minimal memory effects should be developed and used in glovebox-stripper systems.

This paper was prepared in connection with work done under contract number DE-AC09-96SR-18500 with the U.S. Department of Energy. Our colleagues Michael Wood and Davis Nguyen provided the glovebox gas sample data.

1. L. K. Heung et al, "Tritium confinement in a new tritium processing facility at the Savannah River Site", Fusion Technology, vol. 21, Mar. 1992, 594-598.
2. L. K. Heung, "Stripper system performance in the Replacement Tritium Facility", Fusion Technology, vol. 28, Oct. 1995, 859-864.