SUMMARY REPORT ON TASK ORDER NO. C (PHASE II)

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP78-03639A001200110001-9
• Release Decision: RIFPUB
• Original Classification: S
• Document Page Count: 55
• Document Creation Date: December 27, 2016
• Document Release Date: October 25, 2012
• Sequence Number: 1
• Publication Date: December 31, 1958
• Content Type: REPORT

Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
CONFIDENTIAL
SUM NARY REPORT
ON
TASK ORDER NO. C
(PHASE II)
0 `5+r3'~
ATE 26 j UN
j
110" Eft
TYRE
MIMI TV,-.A-,r7 I P 01-1
EAR"G CUSS '5_ PbG t REV CLASS JUST NEXT REV L AUTH: HR 10e2
CONFIDENTI ?
December 31, 1 5
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639A001200110001-9
CONFIDENTIAL
TABLE OF CONTENTS
Page
INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . 1
SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . . . . . . 3
DESIGN OF PROTOTYPE FULL-SCALE GENERATOR . . . . . . . . . . 4
MATERIAL REQUIREMENTS FOR HYDROGEN GENERATION. . . . . . . . 10
Materials List . . . . . . . . . . . . . . . . . . . . . 10
For Generation of Hydrogen . . e . . . . . . . . . 10
For Collection of Hydrogen . . . . . . . . . . . . 11
OPERATING INSTRUCTIONS . . . ... . . . . 11
Procedure 11
Discussion of Operating Procedure . . . . . . . . . . . 16
TECHNICAL PRINCIPLES OF HYDROGEN GENERATION. . . . . . . . . 18
Primary Variables of the Reaction . . . . . . . . . . . 18
Mathematical Expression for the Rate of Reaction. . . . 20
Secondary Variables Affecting Total Reaction Time . . . 22
EXPERIMENTAL PROGRAM . . . . . . . . . . . . . . . . . . 24
Plan of the Experiments 24
Materials . . . . 26
Apparatus . . . . 26
Procedure . . . . 29
Experimental Results. 30
DISCUSSION OF EXPERIMENTAL RESULTS . . . . . . . . . . . . . 35
RECOMMENDATIONS FOR FUTURE WORK . . . . . . . . . . . . . . . 45
APPENDIX I . . . . . . . . . . . . . . . . . . . . . . . . . 47
APPENDIX II . . . . . . . . . . . . . . . . . . . . . . . . . 50
CONFIDENTIAL
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639A001200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639A001200110001-9
CONFIDENTIAL
LIST OF ILLUSTRATIONS
Page
Figure 1. Full-Scale Hydrogen Generator Inflated
for Operation . . . . . . . . . . . . . . . . . . 5
Figure 2. Deflated Hydrogen Generator Packed for Transport
and Inflating Pump. . . . . . . . . . . . . . . 9
Figure 3. Experimental 1/5-Scale Hydrogen Generator . . . . 27
Figure 4. Large-Scale Generation of Hydrogen. . . . . . . . 36
CONFIDENTIAL
71 R~
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639A001200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
SECRET
LIST OF TABLES
Page
Table 1.
Amount of Catalyst Required for Hydrogen
Generation at Temperatures Above 50 F. . . . . . .
15
Table 2.
Hydrogen Generation with Experimental Generator.
in Air . . . . . . . . . . . . . . . . . . . . . .
31
Table 3.
Hydrogen Generation with Experimental Generator
in Water . . . . . . . . . . . . . . . . . .
32
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639A001200110001-9
SUMMARY REPORT
ON
TASK ORDER NO. C
(PHASE II)
December 31, 1958
INTRODUCTION
This report describes the underlying research, and the
design and method of operation of a generator for hydrogen gas
that was developed under Task Order No. C to meet specified
operational requirements and conditions. The research period
covered by this report was from June 28, 1957, through December 31,
1958.
Two reports describing the effort performed under Task
Order No. C were issued previously. A topical report, titled
"An Analysis of the Proposed Hydrogen-Generator Development",
dated September 7, 1956, discussed various means of meeting the
specified requirements for a hydrogen generator. It was necessary
for the generator to produce 3,500 cu ft of hydrogen gas (sea
level) at an isolated site by a controlled reaction in less than
1 hr. Since only surface water would be available at the point
of generation, all chemicals and equipment would have to be trans-
ported, assembled, and/or handled by a single operator; therefore,
the total weight would have to be less than 500 lb, separable into
units of no more than 50 lb each. It was concluded in the analysis
described in the above-mentioned report that a solution of sodium
borohydride could be hydrolyzed to produce the required amount of
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639A001200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639A001200110001-9
hydrogen, and that a generator could be designed to meet the
requirements.
The second report, "Summary Report on Task Order No. C",
dated January 28, 1957, described the results of small-scale
laboratory generation studies, which constituted Phase I of this
research program. The largest of the small-scale tests run pro-
duced 1/100 of the hydrogen required for the desired full-scale
generator. These experiments determined the rate of hydrogen
evolution at various temperatures and involved several different
concentrations of sodium borohydride and catalyst. A number of
different catalysts for controlling the hydrogen evolution were
considered, and cobaltous chloride hexahydrate was studied exten-
sively after it was selected as the preferred catalyst. The
experimental results from that phase of the study established a
preferred method of setting up and carrying through the hydrogen-
generating reaction, to meet the specified operating time of less
than 1 hr over a range of temperatures. It was then possible to
describe a proposed design for a full-scale generator.
It was next necessary to study hydrogen generation on a
larger scale, as described herein, because of other operating
variables introduced as the scale of operation necessarily was
enlarged. As anticipated; an appreciable period of time was
required at the start of the large-scale reaction for adding the
large volume of catalyst solution and for mixing it with the
reaction solution. The rate of loss of reaction heat to the
surroundings was also different in the large generator; this
S E
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639A001200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
changed the pattern of temperature rise in the reaction solution,
and affected the total time of generation.
The experiments described in this report, and identified
as Phase II of this research program, were concluded with one
final full-scale operating demonstration in a prototype generator.
This demonstration was preceded by a number of preparatory 1/10-
and 1/5-scale tests in a 1/5-scale generator. These preparatory
tests were made at less than full scale, to permit reduction of
the over-all cost of the chemicals used in this phase of the
Task Order No. C research program.
SUMMARY AND CONCLUSIONS
The generation of hydrogen in the full-scale demonstration
experiment was completed 26 min after the addition of catalyst
solution to the reaction pool. This was well within the specified
maximum limit of 60 min for hydrogen generation.
Preparations for hydrogen generation were successfully
accomplished by one man, as specified. It appears that it would be
desirable to shorten the time required to prepare the generator
for operation. In particular, the natural flow of water into
the generator through ports below the water line requires about
55 min. Consideration might well be given to shortening the
filling period by providing for and utilizing a more positive
method of filling the unit with water.
L11-
F'r --- T
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
SECREl
The full-scale demonstration was conducted with surface
water at 82 F, which is at the upper end of the anticipated
temperature range. Operation at the lower end of the temperature
range near 32 F would present an additional problem of handling
the increased amount of catalyst necessary to compensate for the
retarding effect of low temperature. The qualitative principles
of modified procedures for operating at low temperature to meet
the specifications are understood. Several small-scale tests
and another full-scale demonstration at low temperature are
recommended to establish quantitatively the effects of the procedural
modifications for operation at low temperatures.
DESIGN OF PROTOTYPE FULL-SCALE
GENERATOR
The hydrogen generator was constructed from neoprene-
coated nylon fabric in the form of a cylinder about 8 ft in
diameter and 30 in. high. The generator was supported by 16
vertical stays inserted in pockets in the cylinder wall. The
stays were held in their vertical position by an inflated double-
walled section of the upper half of the cylinder wall.
Figure 1 is a photograph of the inflated generator
standing on its stays on dry ground. The generator was designed
to operate also while floating in a pool of water. Under this
condition, the inflated air chamber supported the generator on
the water surface, and the stays served to keep the wall section
SEC
El
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
below the chamber extended in its cylindrical form while the
generator was being filled with water and during the period of
gas evolution.
The vertical dimensions of the generator and its air
chamber were so chosen that the amount of water required by the
generator entered through ports below the water line while the
generator was floating free in the water. The two water-entry
ports with open-end fabric sleeves permanently attached are seen
in the foreground of Figure 1. Not visible in the photograph is
a loop at the center of the fabric bottom to which was attached
a weight which pulled the bottom downward during the filling
operation. Also not visible is the valve through which the air
chamber was inflated by a hand pump. Extending above the edge of
the air chamber on the top of the generator was a short open-end
tube, stiffened by a flanged upper edge; this provided access to
the interior of the generator above the water line. This opening
shows in the left foreground of Figure 1 and served several
purposes. After the water entered and was isolated by tying off
the sleeves on the water-entry ports, the solid chemicals required
to make up the generator solution were dumped through the top
opening. The same opening could serve for inspection of the
solution to permit determining whether the added solids were
completely dissolved. The solution could be inspected by dipping
out a small sample and inspecting it for suspended solids, but
this precaution was not usually essential. It was critically
important, however, to measure the solution temperature accurately
S 2 C R_j P'~ T
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
by inserting a thermometer through the opening, because the
temperature determines how much catalyst must be added to generate
the hydrogen in 1 hr or less. Finally, the flanged top opening
served to attach the balloon to the generator. The end of a
flexible tube leading to the balloon was pulled over and tied
below the flanged edge of the open tube before the catalyst
solution was added to the generator.
The catalyst was added as a solution from the catalyst
reservoir, most of which shows in Figure 1 as a collapsed doughnut-
shaped ring of fabric cemented in a concentric position on the
top cover of the generator. The capacity of the reservoir was
about 5 gal, which was sufficient to contain all of the catalyst
solution under most operating conditions. The common wall between
the ring-shaped reservoir and the generator top was pierced by
20 equally spaced 1/8-in.-diameter holes, through which the
catalyst solution flowed and was distributed onto the surface
of the reaction pool. At the right side of Figure 1 is a 3/4-in.-
diameter pressure-equalizing tube bent into a semicircle and
connecting the free space above the reaction pool to the catalyst
reservoir. This acted to prevent the initial pressure, built up
by the generated hydrogen, from retarding the flow of catalyst
solution from the reservoir into the reaction pool.
Lying diagonally across the top of the generator in
Figure 1 is a fabric filling funnel whose flexible spout was
permanently attached to the catalyst reservoir.at a point in the
foreground which is concealed by the projecting air chamber..
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
The upper open-end body of the funnel was large enough to be
pulled downward over the top of a container holding the catalyst
solution. The funnel fabric would be gathered tightly about the
waist of the container, which would then be inverted so as to
pour the catalyst solution through the spout into the catalyst
reservoir. After transfer of the catalyst solution was completed,
the spout was sufficiently long so that it could be tied off to
prevent the escape of generated hydrogen gas.
Careful inspection of the photograph shows four fabric
loops spaced at 90-degree intervals around the top edge of the
generator. These were provided for convenient attachment of ropes
to the generator. The ropes might be necessary in order to
anchor the ballon while it was being filled, or to prevent the
floating generator from drifting while the operator was manipulating
the balloon.
Figure 2 is a photograph of the deflated generator
packed for transport. The package measures 15 x 24 x 35 in.
and weighs 38 lb. The inflating pump is shown held above the
generator to indicate the relative size of these two items.
A drawing of the full-scale generator is included in
this report as Appendix I.
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
Z) t k..0 I I< t= I
SECRET
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
crp1T
E-7.
MATERIAL REQUIREMENTS FOR HYDROGEN GENERATION
The materials and equipment indicated below are required
to generate approximately 3,500 cu ft of hydrogen in the Task
Order No. C generator at an isolated site, as specified, at which
there would be only a pool of surface water sufficiently large
to float the generator. The items of equipment and materials are
listed in the sequence in which they would be needed in order to
carry through the generating procedure.
Materials List
For Generation of Hydrogen
(1) Hydrogen generator
(2) Air pump, capacity about 70 Cu. in. per stroke
(3) Fabric bag for weights, with drawstring, about
300-cu.-in. capacity
(4) Weights, 15 lb, as rocks or sand, probably at site
(5) Rope, i/4 in., about 40 ft
(6) Tools: knife, pliers, dipper, and funnel
(7) Solution container, 5-gal plastic narrow-mouth
bottle with stopper
(8) Sodium hydroxide, 2 lb, in moisture-proof, unbreakable,
nonmetallic package
(9) Rubber gloves, shoulder length
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
F. U I ~ E T
(10) Sodium borohydride, 100--1b of crystals in five
20-lb moisture-proof, unbreakable, nonmetallic
packages
(11) Thermometer, all-metal dial type, 32 to 120 F/1 F
(12) Chart showing weight of catalyst required at various
temperatures of sodium borohydride solution
(13) Cobaltous chloride hexahydrate catalyst, 46 lb of
crystals; in individual moisture-proof, unbreakable,
nonmetallic packages, 3 of 10 lb each, 2 of
5 lb each, 5 of 1 lb each, and 2 of 1/2 lb each.
For Collection of Hydrogen
(1) Balloon and gas connection tube
(2) Tarpaulin
(3) Sand bags
(4) Ropes
(5) Adapter and clamp for connecting balloon to
generator.
OPERATING INSTRUCTIONS
Procedure
The following stepwise procedure was carried through
by one man in one full-scale demonstration run, and is recommended
for field operations:
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
(1) Select a site for operation that provides a water-
pool depth of about 4 ft near the bank.
(2) Unpack the hydrogen generator close to the water
edge. Unfold the generator pack in a position to
favor easy inflation, with the air valve accessible.
(3) Attach the air pump to the air valve. Inflate the
air chamber (with about 700 strokes of the pump)
until the generator stands firmly on its supporting
stays, as shown in Figure 1.
(4) Detach the air pump and close the air valve.
(5) Select about 15 lb of rocks or sand at the site.
Place in the fabric weight bag and tie the mouth
of the bag. The proper weight of rocks may be
estimated by comparison with an equal weight of
packaged catalyst crystals.
(6) Tip the generator and tie the weight bag to the
center of the bottom with the straps provided.
(7) Attach a rope to the generator loop so that the
floating generator does not drift. Anchor the
other end of the rope on shore.
(8)- Slide the generator into the water, and push it
away from the bank until it floats freely.
(9) Push both of the water-inlet tubes into the generator
and lay them along the inside of the generator
wall.
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639A001200110001-9
(10) Tilt and resettle the generator in the water several
times to dislodge any air bubbles trapped beneath
the bottom of the generator. Make certain that
the weight bag is hanging free from the bottom.
Recheck the position of the water-inlet tubes
along the inside wall of the generator as necessary
during the filling period of about 55 min.
(11) While the water is running into the generator,
spread the tarpaulin on the ground, unpack the
balloon, and lay it out on the tarpaulin; be careful
to avoid damage to the balloon fabric. Prepare
the balloon auxiliary equipment for use during
balloon inflation.
(12) After the generator has filled to the proper level,
with the bottom fully extended by the attached
weight, pull out both of the water-inlet tubes and
knot each to shut off the water inlets.
(13) Open the 2-lb package of sodium hydroxide flakes
and pour them into about 1 gal of water in the
5-gal plastic container. Shake briefly to dissolve,
and pour the solution slowly through the open
gas exit into the generator pool. Induce some
mixing by tipping the generator a few times.
(14) Add the sodium borohydride crystals to the generator
pool by opening and emptying one package at a time
into the open gas exit.
r-=r Fd F rte' ('-~ 1
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639A001200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
C v F .,
-14-
(15) Check the pool for undissolved crystals by dipping
out a small sample from the center bottom and
inspecting. Stir with the dipper or tilt the
generator to speed up the mixing and solution, if
necessary.
(16) Measure the temperature of the solution by inserting
the thermometer into generator through the open
gas exit.
(17) Determine weight of the catalyst required by
referring to chart (Table 1).
(18) Make up the catalyst solution in the 5-gal plastic
container. First, fill the container about one-
third full of water, add some of the catalyst, and
shake to dissolve. Add more water (corresponding
to about one third the volume of the 5-gal container)
and the remaining catalyst, and shake again. Fill
the container to the top with more water. Insert
the container stopper, and shake or roll the con-
tainer to mix the solution. Set the container with
the solution upright near the generator at the
water edge.
(19) Attach the balloon to the gas exit of the generator
and arrange the connecting tube so that the balloon
will start to fill without attention while the
catalyst solution is being added to generator.
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
C F. T
TABLE 1. AMOUNT OF CATALYST REQUIRED FOR
HYDROGEN GENERATION AT TEMPERA-
TURES ABOVE 50 F
Weight of Sodium Borohydride: 100 lb
Weight of Sodium Hydroxide: 2 lb
Volume of Catalyst Solution: 5 gal
(made up on basis of values
given below)
Volume of Generator Solution: 560 gal
Generation Time: Approximately 40 min
Initial Temperature Weight of
of Solution, F Catalyst, lb
9o
6
88
6-1/2
85
7-1/2
82
8-1/2
80
9-1/2
78
10-1/2
76
11-1/2
74
13
72
14-1/2
70
16
68
17-1/2
66
19-1/2
64
21-1/2
62
24
60
27
,58
30
56
33-1/2
54
37
52
41-1/2
50
46
SECRET
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
S
(20) Support the full container of catalyst solution
upright on the top edge of the floating generator,
remove the stopper, and pull the open end of the
filling funnel over the neck of the container and
down around the body of the container.
(21) Tip the container and pour the catalyst solution
through the funnel into the catalyst reservoir
on the top of the generator.
(22) Hydrogen generation will start as soon as the
catalyst solution starts to run into the generator
pool from the reservoir. Quickly tie off the
flexible spout of the filling funnel to prevent
any loss of hydrogen.
(23) Attend the balloon during the filling operation.
No further control of hydrogen generation is
needed.
Discussion of Operating Procedure
The foregoing procedure was developed from the experi-
mental results of 19 experiments which measured the time necessary
to generate 1/10 or 1/5 of the required amount of hydrogen in a
1/5-scale generator. One full-scale test was run in the prototype
full-scale generator. With the initial temperature of the generator
pool at 82 F, the addition of 8-1/2 lb of catalyst in 5 gal of
solution resulted in a generation time of 26 min. This was
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
regarded as a good agreement with the generation time of about
40 min predicted from the tests on a smaller scale, inasmuch as
a minimum generation time was desired.
The amounts of catalyst required are given only for
temperatures of 50 to 90 F in the description of the operating
procedure, because the larger amounts which are apparently required
for lower temperatures down to 36 F exceed the solubility of the
solid catalyst in 5 gal of catalyst solution. Also, a larger
volume of catalyst solution would be difficult for one operator
to handle. There are several possibilities for an acceptable
modified operating procedure at low temperatures that would yield
the desired generation time of less than 1 hr. It appears that
the best modification can be selected only after some additional
experimental study is conducted at 1/10 scale in preparation for
a second full-scale experiment.
The type of modifications needed to provide for low-
temperature operation is described later, following a summary
of the technical principles of the hydrogen-generating reaction
in the next section. It may be advisable to include some
discussion of the technical principles in an operating manual
for the generator. Therefore, the principles are described in
the following primarily in a qualitative sense. The quantitative
and mathematical discussion included here could be deleted for
purposes of the operator's manual with only minor changes in
the text.
SECRET
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
S
-18-
TECHNICAL PRINCIPLES OF HYDROGEN GENERATION
Primary Variables of the Reaction
Hydrogen gas is released from a sodium borohydride
(NaBH4) solution by a chemical reaction between the borohydride
ion (BH4) and the water. This reaction does not occur unless
there is a catalyst present in the solution. After the catalyst
is uniformly mixed in the solution, the rate at which hydrogen
is generated depends primarily on three variables:
(1) the temperature of the solution
(2) the concentration of the catalyst
(3) the concentration of the borohydride.
It is the combined effect of these variables that determines
the speed of the reaction at any moment, and thus the time required
for the reaction to be completed.
The three variables listed above differ in their
individual effects on the rate of hydrogen generation. The
generation rate increases with rising temperature, and doubles
for a rise of about 15 F. The generation rate also increases in
proportion to an increase in the catalyst concentration, so that
doubling the concentration of the catalyst would double the genera-
tion rate, if both the temperature and the borohydride concentration
remained the same. Finally, the borohydride concentration has the
least effect on the generation rate. Increasing the borohydride
concentration by almost 50 per cent reduces the total generation
time by less than 10 per cent.
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
SECRET
During the time that hydrogen is being generated, the
temperature and the borohydride concentration are continuously
changing, and the hydrogen generation rate changes also. The
reaction which produces hydrogen releases heat and consumes the
borohydride. Most of the heat remains in the reacting solution
and serves to raise its temperature, although some heat is lost
continually to the surrounding water and to the air outside the
generator.
The total reaction time is thus greatly affected by
the initial temperature of the solution and by the amount of
water in the generator. Both influence strongly the range of
temperature that is covered during the reaction, and must be
carefully determined in order to control the reaction time. If
the amount of water added to the generator is intentionally
reduced, then a much shorter generation time is possible. The
concentrations of borohydride and catalyst are increased by
dissolving fixed weights of these solids in less water; also, the
temperature rise occurring during the generation reaction is
larger in proportion to the reduction in the volume of the
solution. This occurs because, regardless of the amount of
water used, approximately the same amount of heat is released
in the complete reaction, and it heats the smaller amount of
solution to a higher temperature. For example, the maximum temper-
ature rise for 100 lb of borohydride in 560 gal of water is 58 F
above the initial temperature of the solution. Using only 350 gal
of water would lead to a temperature rise of about 93 F above the
initial temperature.
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
SECRET
Mathematical Expression for the Rate of Reaction
According to the conventional methods of expressing
reaction rates, the following rate equation was derived from the
experimental data obtained in isothermal bench-scale tests, and
was used to calculate the qualitative comparisons in the preceding
section:
- d (NaBH4) = 1.45 x 1013 (CoC12)(NaBH4)0.39e-179500/RT
d (NaBH4) = the differential rate of consumption of
dt
NaBH4 in moles per (liter)(second) at
time t in the assumed reaction:
NaBH4 + 2H20
catalyst
NaBO2 + 4H2
(NaBH4) and (CoC12) = the concentrations, in moles per
liter, of the compounds in the
reaction solution at time t
1.45 x 1013 = the value of the rate constant*
for the catalyzed reaction
* The rate constant is conventionally designated as k in reaction-
rate equations. The value given here was calculated from the bench-
scale tests. The numerical value of k appears, however, to vary with
the scale of the test. It is probable that certain secondary variables
affect the over-all reaction rate as the scale of the reaction is
changed from bench scale to 1/5 scale to full scale. It was found
that more catalyst than would be predicted from this equation was
needed in order to attain the desired reaction time in the larger
scale tests. If the results of these tests are used to calculate
the rate constant, its value for the full-scale reaction is estimated
as 3.6 x 1012.
SEcR E
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
SECRET
e-17:500/RT = the dimensionless Arrhenius constant,
aE
e RT , which expresses the varia-
tion of the reaction rate with
temperature. When the activation
energy, AE, is expressed as 17,500
calories, the gas constant, R, is
2.00 calories per degree K, the
temperature T is in degrees Kelvin,
and e is the Naperian logarithm
base 2.718.
The rate equation is based on the usual assumptions that
temperature and the concentrations of the reactants are uniform
throughout the reacting solution. To calculate the total reaction
time, the integrated form of the equation is first used to calculate
the approximately constant rate over a short time interval of
arbitrary length, such as 1 or 5 min. The calculated rate of
consumption of NaBH4 and the length of the time interval are used
to calculate the amount of NaBH4 reacting during the interval
(i.e., rate x time = amount). The corresponding temperature rise
is then calculated from the heat liberated by the amount reacted.
The temperature rise and NaBH4 consumption as determined give new
values for the solution temperature and NaBH4 concentration at the
end of the time interval. These are substituted into the rate
equation for the calculation of the rate during the next interval.
When this stepwise procedure shows that the NaBH4 has been exhausted,
a summation of the time intervals gives an approximate value for
the total reaction time.
SECRET
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
SECRE1
Secondary Variables Affecting Total Reaction Time
When a large reactor is used to generate the considerable
quantities of hydrogen required, an appreciable time elapses before
a uniformly reacting system is established that corresponds
approximately to the ideal system discussed above. The quantitative
details of the events occurring at the beginning of the reaction
are not completely understood. However, two secondary variables
are believed to have a major influence on the total reaction time.
These are (1) the time required for the formation of the active
catalyst after the addition of the cobalt chloride solution to
the reactor pool, and (2) the time required for the dispersion of
the catalyst throughout the reactor pool so that it can initiate
hydrogen evolution.
It is known that the active catalyst is a solid product
of a preliminary reaction of the cobalt chloride catalyst solution
as it is added to the sodium borohydride solution. The preliminary
reaction occurs rapidly, with the evolution of heat and hydrogen
and with the consumption of some of the sodium borohydride. When
larger amounts of cobalt chloride catalyst are added to counteract
the slowness of the catalyzed reaction at low temperatures, the
preliminary reaction may consume an appreciable part of the sodium
borohydride. In this case, although the total evolution of hydro-
gen is affected only slightly, the initial concentrations in the
system and the corresponding reaction time may be significantly
changed.
S, F
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
SECRET
The effective concentration of solid catalyst, which
controls the rate of hydrogen evolution, may be affected not only
by the total weight of cobalt chloride added,but also by the average
particle size and the solid surface area per unit weight. The
properties of the solid active-catalyst particles may be strongly
influenced by the temperature at which precipitation occurs, by
the concentration of the cobalt chloride in the solution, and by
the rate of addition.
The time required to disperse the active catalyst uniformly
throughout the reaction pool after it is precipitated may be a
significant part of the total reaction time. The active-catalyst
particles formed at the top of the reactor pool fall toward the
bottom by gravity, but the rate of fall is slow for these small
particles. The downward motion is opposed by the rising bubbles
of hydrogen, that are evolved exclusively at the solid surfaces,
and also by the upward currents of the solution heated by the
reaction.
The catalyst solution is distributed over the surface
of the borohydride-solution pool from multiple holes in the catalyst
reservoir so that the horizontal distance to the center and edges
is minimized. Agitation by the hydrogen bubbles would tend to
cause circulation and horizontal dispersal of the particles after
some of them reached the bottom.
Although the qualitative aspects of the complex variables
which affect the length of the initial time period are deduced from
a few experiments, times for large-scale reactions cannot be predicted
I v CT
S S C R
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
-24-
precisely by extrapolation. The only alternative is to conduct
experiments over the total range of conditions which may be met
in field operations. Therefore, an additional program of tests
at low temperatures has been recommended, as discussed later
in this report.
EXPERIMENTAL PROGRAM
Plan of the Experiments
It was pointed out previously that the rate of the
catalyzed hydrolysis of NaBH4 was originally studied in laboratory-
scale tests (Phase I). The time required for the complete reaction
to occur can be adjusted by adding a selected amount of catalyst.
A reaction rate equation was deduced from these experiments that
was presented earlier in this report. The equation, as written,
describes the reaction rate at a fixed temperature and at fixed
concentrations of reactants. However, the form of the equation
shows that, to maintain constant total reaction time at different
temperatures, the logarithm of the weight of the catalyst must be
varied directly as the reciprocal of the initial temperature of
the reacting solution; it is also evident that all other reaction
conditions must be the same.
In connection with this type of problem, the ideal of
exact similarity cannot be reached even in replicate experiments.
A change in the scale of the experiment introduces larger changes
in the uncontrolled secondary variables, such as the time for the
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
C R, E T
-25-
formation and mixing of the active catalyst in the reaction pool,
and these affect the total reaction time. The quantitative effect
of these uncontrolled variables had to be measured experimentally,
and this was the general objective of the experiments described
in this report. It was also the purpose of these larger scale
experiments, to provide bases for developing a procedure which
would lead to reproducible results in the field, and for examining
the generator design for possible shortcomings which could be
eliminated in a full-scale prototype unit.
It was impossible to investigate each operating variable
individually, because this would have required an excessively
large number of experimental runs. Therefore, the first group of
runs was designed to explore the combined effects,on the reaction
time and completeness of reaction, of the concentration of the
reactants, temperature, method of catalyst addition, volume of
reaction solution, and generator ,design. Another group of runs
studied specifically the effect of the amount of catalyst on the
reaction time, and a third, the effect of the initial temperature
on the reaction time. Last, the possible effect of natural surface
water on the reaction time was measured in preparation for the
final full-scale demonstration run. The full-scale generator was
also tested in the field to determine the best procedure for
filling it with water and for preparing the reaction pool.
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
-26-
S E C R tF,-- "' I
Materials
Sodium borohydride was supplied by the Sponsor, and was
a product of Metal Hydrides, Inc., of Beverly, Massachusetts. The
manufacturer's specifications indicated 98 per cent purity for
the material provided. Analyses of three samples gave assays of
93, 95, and 98 per cent NaBHk. The cobalt chloride catalyst was
cobaltous chloride hexahydrate (purified, crystals) obtained from
Matheson, Coleman, and Bell. The sodium hydroxide used was Baker &
Adamson chip sodium hydroxide (97 per cent min). River water was
generally used in making up the solutions. (unless indicated
otherwise).
Apparatus
Figure 3 shows the experimental 1/5-scale generator as
finally modified during the course of these experiments; a
drawing of this unit is included in this report as Appendix II.
The 1/5-scale generator was essentially the same as the full-scale
generator already described, except that the diameter was smaller,
so as to provide 1/5 the volume of reacting solution with the same
solution-pool depth. The ring-shaped catalyst reservoir was
cemented on the inside of the top; thus, it does not show clearly
in Figure 3. The same experimental generator was used for the
1/10-scale tests by filling it with solution only to one-half
the normal depth (normal depth, 18 in.).
SECE
ET
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639A001200110001-9
SECRET.
-27-
FIGURE 3. EXPERIMENTAL 1/5-SCALE HYDROGEN GENERATOR
SECRET
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639A001200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
This experimental generator was approximately 43 in.
in diameter and 33 in. high, and was made from neoprene-coated
nylon fabric approximately 0.020 in. thick. The unit was supported
by wooden stays and by an inflated ring of 3-in. annular thickness
that was attached to the top and was inflated with the use of a
hand pump. Access ports for the addition of the catalyst solution,
for the entry of the water, and for inspection purposes were
provided by three gas-tight brass zippers. This experimental
generator was designed to withstand 0.4 psig and a maximum
temperature of 165 F.
The gas outlet was fitted with a brass collar holding
two thermocouple wells and an exit tube for the generated hydrogen.
In the course of the experiments, the evolved gases were passed
through a meter before being exhausted to the atmosphere by a
specially constructed exhaust system, which minimized the fire
and explosion hazards. The temperature of the reacting solution
was measured by single-junction iron-Constantan thermocouples
inserted into the generator through the gas outlet. Barometric
pressures were read from either an aneroid or a mercurial barometer.
During the initial series of tests, leakage occurred
around the zippers, which were difficult to operate and eventually
were broken. It was concluded that the zippers would be unreliable
for field operation of such a unit.
The zippered water inlet at the bottom edge, and the
zippered catalyst-solution entry port in the top were patched and
sealed. A 4-in.-diameter neoprene-coated fabric water-inlet tube
S Fk=1 C R' E
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
and drain was subsequently fabricated and assembled to the catalyst-
solution ring. A neoprene-coated fabric bag was attached to the
end of the catalyst entry tube. This bag was large enough so that
it could be pulled over a 5-gal bucket; thus, to make the catalyst-
solution addition, the bucket could be up-ended so as to simulate
the contemplated procedure of pouring the catalyst solution into
the full-scale generator. This experimental generator, so modified,
was used beginning with Run 9 and the modifications were found
to be quite satisfactory.
Procedure
In a standard 1/5-scale experiment, the inflated
1/5-scale generator was filled with 110 gal of water pumped from
two 55-gal drums, and a weighed amount of NaBH4 (approximately
20 lb) was dissolved in the water. The temperature of the pool
was read from a thermometer dipped in the solution. An appropriate
weight of catalyst, determined on the basis of the measured
temperature and the desired total generation time, was dissolved
in 2 gal or less of water and added to the generator pool. At
first, the catalyst weight to be used was read from a chart of
catalyst weights and temperatures given in the earlier report on
the Phase I effort. The results obtained from the early experiments
described herein quickly showed that these catalyst weights were
insufficient to produce complete reaction within the 60-min limit
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
SECRET
for large-scale tests; consequently,revised values for catalyst
weight were deduced from the experimental results and utilized.
(The revised catalyst weights were given in Table 1.)
At times, the standard amounts of water and borohydride
were varied somewhat to accommodate the objectives of individual
experiments. Also, the catalyst solution'was, on occasion, added
through a funnel with a long curved spout, which could discharge
the solution at a selected location near the edge or at the center
of the pool, instead of distributing it over the top of the pool
from the perforated ring.
It had been shown in the earliest tests that the measured
volume of hydrogen corresponded consistently to almost complete
reaction, and that the course of the reaction could be followed
by noting the rise in the temperature of the reaction pool. There-
fore, in the later tests, the gas meter was by-passed, and the time
when the temperature of the reaction pool stopped rising; was
measured as indicating complete reaction.
Experimental Results
Table 2 gives the experimental results of the 1/10- and
1/5-scale runs in the 1/5-scale generator in air; Table 3 lists
similar information for the runs in which the reaction pool within
the experimental generator was surrounded by water on the outside.
For each table, Column (1) gives the run numbers, which were
assigned in chronological order, and Column (2) indicates the
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
SECRET
Vol of Hydrogen
(Dry,STP), cu ft
Temperature,
F
Catalyst Added
(CoC12.6H20)
Total
Reaction
Run
Theore-
Experi-.
Amount,
Time,
No.
tical
mental Initial
Rise
lb.
Method
min
Remarks
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
360
--
53
52
0.92
From ring
' 65
Preliminary
2
360
--
62
--
0.71
From ring
70
trial run
Preliminary
3
720
686
62
63
1.37
From ring
85
trial run
Slow rate
4
360
--
61
53
0.96
At center
58
H2 leak
5
648
599
54
46
1.62
At edge
150
Reduced
6
772
720
46
61
2.02
At center
125
NaBH4
concentra-
tion
7
772
744
64
74
1.83
At center
70
Increased
8
772
764
55
47
2.02
At center
125
NaBH4
concentra-
tion
Reduced
9
720
686
66
46
1.67
From ring
64
NaBH4
concentra-
tion
110
720
--
63
57
2.66
From ring
59
Note: The numbers in parentheses, located under the last line of each of the
column headings, are provided to facilitate the presentation and com-
prehension of the discussion of the research in the text.
SECRET
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
SECRET
Vol of Hydrogen Temperature, Catalyst Added. Total
(Dry,STP), cu ft F (CoC12.6H20) Reaction
Run Theore- Experi- Amount, Time,
No. tical mental Initial* Rise lb Method min
(1) (2) (3) (4) (5) (6) (7) (8)
11 360 364
12 360 373
13 360 354
14 360
15 360
16 720 a
17 720
18 370
19 370
20** 3,700
47 37 5.58 From ring 34
47 29 4.18 From ring 59
47 33 5.15 From ring 41
66 35 1.98 From ring 42
65 29 2.17 From ring 35
66 46 4.41 From ring 23
55 57 11.70 From ring ._10
85 23 0.91 From ring 30
84 33 0.78 From ring 37
82 55 8.60 From ring 26
Remarks
(9)
Pond water
used
Pond water
used
Pond water
used
? Note: The numbers in parentheses, located under the last line of each of the
column headings, are provided to facilitate the presentation and com-
prehension of the discussion of the research in the text.
* The surrounding water was also at this temperature.
This was the full-scale demonstration run in the. prototype generator.
SECRET
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
SECRET
theoretical amount of hydrogen corresponding to the weight of NaBH4
used in each run. In the calculation of the values in the table
from the weight of NaBH4 used, it was assumed in each case that
the NaBH4 was 95 per cent pure. Some runs show target volumes of
hydrogen of 720 cu ft and others, 772 cu ft, but these are all
nominally 1/5-scale runs. These differences arose from slightly
different assumptions as to the NaBH4 purity, in the calculation
of the amount to be used in the experimental runs. It should be
recognized that these differences had no significant effect on
the results obtained and the conclusions drawn. The same explanation
applies to the nominally 1/10-scale runs, which, as described
in the tables, show target hydrogen volumes of 360 and 370 cu ft.
Column (3) shows the standardized volume of hydrogen
generated experimentally for those runs in which it was measured.
The values given were calculated from the measured volume, tempera-
ture, and pressure of the metered gas, assuming that the gas
was saturated with water vapor and was all at the measured temperature.
The gas temperature rose continuously in the course of
each run, so that the temperature and water-vapor corrections had
to be calculated for incremental volumes in order to obtain by
summation the total correction. Considering also that the actual
purity of the NaBH4 undoubtedly varied from run to run, the agreement
between Columns (2) and (3) is considered to be remarkably good.
The ratio of the values in Columns (3) and (2), representing the
completeness of the reaction, varied between 93 and 103. per cent.
Following the performance of Run 13, the volume of hydrogen
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
U ,_ II
-34-
evolved was not measured, since it was then known that the reaction
went to substantial completeness.
Columns (4) and (5) give the initial temperature of
the reaction pool and the temperature rise noted at the end of
the reaction. In the runs described in Table 3, the experimental
generator wall was cooled by the surrounding water which was at
the initial temperature; the temperature of the surrounding water
did not change appreciably during any of the runs.
The temperature rise for this reaction using the standard
concentrations* would be 58 F without heat losses through the
walls of the generator or in the exit gas and vapor. In most of
the runs, the temperature rise generally followed a qualitatively
appropriate pattern. Runs 5 and 8, at reduced concentration,
showed a temperature rise below normal, and Run 7, with a more
concentrated solution, yielded a higher than normal temperature
rise. In Run 5, the solution volume was normal and less NaBH4 was
used. Run 8 involved a reduced concentration which was obtained
by adding 15 extra gal of water to the pool. In Run 7, there was
a deficiency of 10 gal of water.
The water-cooled runs in Table 3 generally showed, as
expected, less of a temperature rise than did those described in
Table 2 that were air cooled. Within Table 3, the data on temper-
ature rise showed the expected correlation with the duration of
* For 1/5 scale, these are 110 cu ft of water and approximately
20 lb of NaBH4.
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639A001200110001-9
the run fin Column (8)J; the, shorter the total reaction time,
the less was the opportunity for heat losses to occur through the
generator walls, and consequently, the greater was the increase
in temperature.
Column (6) indicates the weight of solid catalyst added
in water solution to the generator in each case, and Column (7),
the method of addition of the catalyst solution. Column (8)
lists the time interval between the start of the catalyst-solution
addition and the end of hydrogen evolution, i.e., the total time
for the hydrogen-generation reaction.
DISCUSSION OF EXPERIMENTAL RESULTS
Figure 4 plots for each experimental run the catalyst
weight against the initial temperature. The weight scale is
logarithmic; the temperature scale is linear with the reciprocal
of the absolute temperature. Catalyzed reactions requiring the
same reaction time at different temperatures should theoretically
fall on a straight line when the catalyst weight and the initial
temperature are plotted on such scales, if all other reaction
conditions are comparable.
Two straight lines are shown on Figure 4. The upper
line is discussed later. The lower line was constructed from the
experimental results of the previously conducted 1/100-scale
tests, (described in the "Summary Report on Task Order No. C"
dated January 28, 1957), and represents the amount of catalyst
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639A001200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
%J"%1sikN11 . 1
-36-
40
0
U-
L 2
v 9
o 8
100
90
80
70
3
90
/
---
--- - 1
L7
V
34
11
-
--
-
13 41
--
-
-
12 59
O
a,
2
L
15 35
~
~~
1442
If
9
~
(45O
L0\
h
Q25
25
4
0 6
ij
4
70 2
-B
5
^ Tenth scale
Fifth scale
0 Full scale
Run numbers inside points
Time (in minutes) in i/o/ics
80 70 60 50 40
Temperature, F
FIGURE 4. LARGE-SCALE GENERATION OF HYDROGEN
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
SECS
necessary to bring about a complete reaction in 60 min at these
temperatures in the laboratory-scale runs. However, all of the
experimental points shown on Figure 4 are for large-scale runs,
described in this report; the run number is inserted inside each
point, and the time in minutes for complete reaction is shown in
italics at the side.
In the interpretation of the significance of the experi-
mental results, the individual experimental points in Figure 4
are considered in terms of the corresponding experimental time
periods as shown. If reference is made to the first series of
10 runs (Table 2), it should be noted that all of the corresponding
points would have fallen directly on the lower line of Figure 4
if the catalyst weights had been so chosen, at the option of the
experimenter. Only Runs 2 and 3 used exactly the weight of catalyst
called for by the 60-min line for the 1/100-scale tests. Further,
in these two runs, the times for complete reaction were appreciably
longer than 60 min. It thus appeared that the total reaction time
was longer in the large-scale runs because of reaction variables
which differed from those prevailing in the laboratory-scale
reactions. The remainder of the series are grouped around the
60-min line, and represent departures from the standard conditions
as outlined specifically in Table 1. The three points which fall
below the lower line represented runs in which considerably less
than the standard amounts of catalyst* were used; this in itself
* It should be emphasized that any discussion of the amount of catalyst
used involves directly a consideration of the initial temperature of
the reaction pool. Thus, in any comparative review of the data in
Tables 2 and 3, and particularly the data on the amount of catalyst
used, it should be noted that the initial temperatures varied considerably.
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
-38-
would be expected to retard the reaction and thus increase the
time. In Run 5, the concentration of NaBH4 was lower, and the
catalyst solution was added at the outside edge of the pool at
one location. Either of these two variations alone would also
be expected to increase the reaction time. All of these factors
in combination caused a large increase to 150 min in Run 5. Run
6 involved a less critical addition of catalyst at the center of
the pool, and the 125-min reaction time obtained was slightly
less unfavorable.
In Run 7, the retarding effect of pouring the catalyst
solution into the center was partially compensated for by increasing
the amount of catalyst and increasing the concentration of NaBH4;
but, the resulting reaction time of 70 min was still above the
value predicted from the laboratory-scale tests. It was concluded,
however, that the retarding effect of the center pour could not
account entirely for the relatively long 70-min reaction; and
that unidentified scaling effects were also involved. Run 8
confirmed qualitatively the retarding effects of low NaBH4
concentration and center pour, partially counteracted by an
increased amount of catalyst.
Runs 9 and 10 completed the first series of runs; these
involved successively increased amounts of catalyst with other
conditions maintained standard. Run 10, with a total reaction
time of 59 min, fell barely within the 60-min maximum; the amount
of catalyst used was about twice that shown by the lower line in
Figure 4.
SECRET
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
SECRET
-39-
It thus became apparent that there was a definite effect
encountered in scaling the reaction to a larger size and this
increased the total reaction time considerably. Although this
scaling effect is not completely understood and was referred to
earlier as "unidentified", the evidence points to inefficient
mixing of the catalyst solution vertically in the generator
(1/5 scale) unit. This retarding effect might be overcome by
injecting the catalyst solution at the bottom of the reaction
pool, so that natural convection induced by the temperature rise
and by the vigorous gas evolution would cause more positive
mixing, and would reduce or eliminate the probability of a non-
reacting bottom layer during the early part of each run. A redesign
of the system to provide for this method of catalyst-solution
addition, and a corresponding study of a new associated procedure
did not appear as advisable, over all, as the simpler alternative
of increasing the amount of catalyst used, in order to counteract
the slower reaction characteristic of the larger scale generation
operation.
Therefore, a second series of runs, Nos. 11 through 18,
was carried out to establish a new reference curve for catalyst
requirements. It was known qualitatively that the cooling effect
of water surrounding the generator would also retard the reaction,
so, in this series, the generator was floated in water. It was
estimated on the basis of detailed calculations that a reduction
in reaction time by about one-third would probably be attained
by doubling the catalyst weight, under otherwise equivalent
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
SLKE I
conditions. Also, it was decided that it would be worth while to
work toward a target time of about 35 to 40 min, to insure completion
of the generation operation well within the specified time limit.
It was felt that a 40-min target time would provide a safe margin
for possible variations in the scaling effect, which was not
completely understood and so could not be positively controlled.
Also, the shortest possible generation time was desirable in itself.
On the basis of these considerations, Runs 11, 12, and 13
were made at 47 F, and these successfully bracketed the target
time of 40 min, as anticipated. Another group of three runs at
about 66 F again bracketed the target time, and provided the basis
for the construction of the upper line of Figure 4, which shows
the amount of catalyst required to complete large-scale generations
in 35 to 40 min. Run 17 was partially successful in establishing
a safe lower time limit; the completion of the reaction was timed
at about 10 min. A large increase in the amount of catalyst used,
in excess of the corresponding amount indicated by the new reference
line (the upper line of Figure 4), showed that the reaction could
operate predictably and fairly smoothly in much shorter times.
However, in Run 17, the capacity of the gas exit was exceeded,
and the experimental fabric generator was damaged slightly.
The 1/5-scale generator was subsequently repaired and
moved to a pond of surface water, where it was used in conducting
Runs 18 and 19, to examine the possible effects of higher initial
temperature and of the impurities of the natural water on the
reaction time. As expected, no significant effect of water impurities
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
SECRE I
was found. However, there appeared to be some trend toward shorter
reaction times at higher temperatures.
The experimental program was concluded with a full-scale
demonstration, Run 20, in the prototype full-scale generator,
following the procedure outlined in detail in the early sections
of this report. The reaction was successfully completed in
26 min without damage to the generator. A failure in an "on the
spot" improvised fitting connecting the generator gas outlet to
the balloon tube prevented complete filling of the test balloon;
however, this was not attributable to the generator design.
A trend exists at high initial temperatures toward
generation times which are shorter than those predicted by the
upper line of Figure 4. This is satisfactory and even desirable.
The same trend implies an undesirably large increase in the genera-
tion time at the extreme low end of the anticipated-service
temperature range; it appears that some adjustment in operating
procedure will have to be made to counteract this trend.
The trend to longer generation times at the low temperatures
is also indicated by the greater slope of the upper line in Figure 4
as compared to the lower one. Thus, the upper line shows proportion-
ately larger catalyst requirements at the low temperatures. It
is believed that this trend is qualitatively distinct from the
scaling effect attributed to stratification and inefficient mixing,
which have previously'been discussed as the reason for, a portion*
* The remainder of the upward displacement was attributable to the
decision to reduce the target total generation time from 60 min to
35-40 min.
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
of the upward displacement of the large-scale reference line in
Figure 4. It seems probable that there exists a catalyst concentration
effect that has not been accounted for in the basic reaction
equation. Because the active catalyst is a solid formed from the
cobalt chloride solution, uniform mixing would produce a cloud of
active-catalyst particles in the reaction pool, and each particle
would be reacting within its own sphere of surrounding solution.
The greater the number of active-catalyst particles, the smaller
the sphere of reaction of each, or the greater the tendency for a
"blanketing" or interference of each particle with its neighbors.
Thus, with increasing amounts of catalyst used, and this is the
case at the lower temperatures, the activity of the catalyst per
unit weight would be reduced, with the resulting trend toward
increasing generation time at the low temperatures.
It is not possible to increase the amount of catalyst
sufficiently to counteract the increase in total generation time
at the low temperatures, because of the limited solubility of the
cobalt chloride catalyst in water at such temperatures. The present
procedure calls for dissolving the catalyst required for a full-
scale run in 5 gal of water; this represents about the maximum
volume and weight of catalyst solution which can be handled by
one man when the solution is in a single container. At temperatures
of less than about 45 to 50 F, the solubility of the cobalt chloride
catalyst in 5 gal of water is about 45 lb per 5 gal of solution,
or less than the catalyst requirements shown on the upper line of
Figure 4 for the temperature range below 50 F.
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
SECRET
-43-
In addition to the limitation on cobalt chloride solubility..
at the low temperatures, it may be undesirable to add excessively
large amounts of catalyst because of the rapid evolution of hydrogen
during the preliminary reaction which forms the solid active-catalyst
particles from the cobalt chloride solution. Based on calculations
made early in this program, 10.6 cu ft of hydrogen are released '
in the initial surge for each pound of catalyst added. The initial
surge of gas might reach dangerous proportions, although it could
probably be controlled by reducing the rate of flow of the cobalt
chloride solution into the pool.
To provide for generation at the low temperatures, three
variations in the procedure to overcome solubility limitations
might be made without necessitating a major change in the full-
scale generator design:
(1) The amount of catalyst might be reduced to give a
longer generation time of about 55 min.
(2) The amount of cobalt chloride that could be dissolved
might be increased by heating the catalyst solution,
by means of auxiliary equipment which might be
provided within the over-all total weight limitation.
(3) The volume of water in the reaction pool might be
reduced for low-temperature generation, without
changing the quantity of NaBH4 used; thus, the heat
of reaction would raise the reaction pool to a higher
temperature, which would accelerate the reaction,
thereby shortening the generation time or reducing
the amount of catalyst required.
These variations might be utilized Aigly or in combination.
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
L
Experimental runs would be necessary in order to measure
the quantitative effects of these three variations. The general
objective of such experiments would be to develop the simplest
procedure that would permit satisfying the reaction specifications
at low temperatures.
Although it would be preferable not to make any change
in the full-scale generator design, the following suggestions
requiring both a design and procedural change are presented in
order to complete this discussion; both changes would be directed
toward achieving more efficient, mixing of the catalyst solution
in the reaction pool.
It is deduced, from comparison of the effects of pouring
the catalyst solution in through the ring and at the center of
the pool, that lateral distribution occurs more readily than was
originally expected. It, therefore, seems probable that more
efficient mixing could be attained by adding the catalyst solution
through a tube to the bottom center of the reaction pool. This
should induce a circulation pattern upward at the center and
downward around the generator wall and should rapidly spread the
catalyst to all parts of the pool. However, some positive method
of forcing the catalyst solution through such a tube would probably
be required to counteract the back pressure which would build up
as soon as the reaction started.
The other change which might be applicable would be to
make the catalyst addition by dumping pelletized solid anhydrous
cobalt chloride into the pool. The pellets would presumably fall
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
ET
-45-
by gravity to the bottom before they dissolved. There would be
no problem of ultimate solubility in the whole pool, as there is
for a limited amount of catalyst solution made up before addition
to the pool. Material handling would be easier with the pellets,
as the weight and volume of catalyst required would be much lower
than the total weight and volume of 5 gal of solution. The maximum
weight of catalyst might be about 25 lb for the lowest temperature,
and this amount, in the form of pellets, would occupy a volume
of perhaps 2 gal or less.
However, the addition of solid catalyst would require:
(1) a different type of catalyst reservoir which could be filled
with the pelletized catalyst, closed, and then manipulated to
permit dumping the solid into the generator pool; (2) formulation
of a catalyst pellet that would sink to the bottom and then
disintegrate to a fine powder so as to facilitate rapid dissolution;
and (3) a series of experimental runs to confirm these predictions
and establish the details of a revised procedure.
RECOMMENDATIONS FOR FUTURE WORK
It is recommended that additional work be done to provide
for hydrogen generation in the prototype full-scale generator at
low temperatures. It is suggested that such future work should
be directed toward a study of the reduction in catalyst requirements
at low temperatures by working to a target value of 55-60 min for
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
SECRET
the total generation time, and by reducing the amount of water
used in the reaction pool so as to increase the associated tempera-
ture rise when the hydrogen-generation reaction occurs.
SECRET
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
t~ 1= U< I= I
-47-
Drawing of Full-Scale Generator
SECRET
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639A001200110001-9
-49-
CONFIDENTIAL
GOODYEAR "SWIMMING TANK" REPAIR KIT
The repair kit provided by the manufacturer of the
experimental generators consisted of the following:
Quantity
Material
Goodyear No.
4 1/2-pint cans of patching cement
1451-C
4 8-cc bottles of cement accelerat
or
983-C
1 1-inch cement brush
--
1 2-1/4 x 2-1/4-inch roller
--
24 1-inch-diameter patches
--
1 Linear yard of neoprene-coated n
ylon
--
fabric
1 Instruction booklet
--
Q FIDE TIAL
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639A001200110001-9
 Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9
L
-50-
Drawing of Experimental 1/5-Scale Generator
SECRET
Declassified and Approved For Release 2012/10/25: CIA-RDP78-03639AO01200110001-9