position of source and detector 
o6 Intersection- 31.3 cm 
~—By thermistor 31.3 cm 
FSD AA aaa 
IS 2 
Gounting Rate (cps) 
FIG. 5. Source is stationary at 15-cm above 
Low rise in counting rate as de- 
tector moves past liquid level is undesirable 
liquid level. 
passed, the rise in count, although not 
as great as before, is quite sharp. 
Behavior similar to that shown on 
curve C is expected if the source were 
to be fixed outside the tank relatively 
far below the water level, with the 
region around the level being scanned 
by the counter. This is confirmed by 
Fig. 4, curve A, for which the fast- 
neutron source was fixed to the outside 
of the tank 15 cm below the level. 
Curve B in this figure shows the results 
when a block of paraffin wax, shaped to 
fit symmetrically over the source, was 
placed against the tank, covering the 
source with approximately 6 cm of wax. 
The paraffin produced essentially a 
slow-neutron source of greater inten- 
sity, but otherwise the curve obtained 
was similar to that with the fast-neu- 
tron source. Principle advantage of 
converting a fast-neutron source to a 
slow one appears to be that less radio- 
active material is needed. 
Since the slow neutrons in the liquid 
inside the tank largely determine the 
counting-rate curve shape, an equally 
sharp break at the liquid level is ex- 
pected on placing the stationary source 
inside the liquid rather than outside of 
it. This was confirmed by experiment. 
Source above counter. The previ- 
ous results demonstrate that when the 
source is positioned below the counter, 
it should not be placed too close to it. 
Experiments with the source above the 
counter indicate that the source should 
not be too far away in this case. With 
the center of the source as close as 2.0 
5 
§ 
: 
3 
Liquid level by 
Intersection | Thermistor 
32.7cm 326 cm 
31.7 cm 31.95cm 
FIG. 6, 
15-cm below the level. 
tor, 8.3-cm above it. 
em above the counting element, the 
counting curves showed sharp discon- 
tinuities at the liquid level and resem- 
bled those in Fig. 4. Figure 5 presents 
the case where the source is stationary 
and relatively far (15 cm) away from 
the liquid level—note the minimum 
corresponding to the liquid level. The 
important point demonstrated here is 
that the rise in counting rate as the 
counter passes down under the level is 
relatively small. This undesirable fea- 
ture would be expected to occur also 
when the source is moved with the 
counter but fixed at a relatively large 
distance above it. 
The previous results were obtained 
with a brass container having a wall 
thickness of only 0.16 cm (1746 in.). It 
was expected that a thick container 
wall would tend to make the discon- 
tinuity in neutron count at the liquid 
levellesssharp. There are two reasons 
for this: the detector does not pass as 
closely to the liquid-gas interface; and 
the neutrons are scattered within the 
container wall. Figure 6 presents data 
obtained when the tank-wall thickness 
was increased by placing a steel sheet 
1.27-cm (14-in.) thick against the brass 
wall. The scintillation counter was 
used in these experiments. Curve A 
was obtained with the source inside the 
liquid, 15 em below the level and 15 cm 
inside the tank wall. The break in the 
curve is sharp and indicates the level 
accurately. With the source 8.3. cm 
above the counter, curve B in this 
figure, the break in the curve is not as 
Counting Rote (cps) 
For A, source was inside the liquid, stationary at 
For B, source moved with detec- 
Thicker-walled tank was used 
well defined. These examples illus- 
trate the need for determining the best, 
relative location of the source for a 
given thick-walled container. For 
thin-walled containers there is a rather 
wide choice of source locations. 
Conclusions 
The level of a hydrogenous liquid in 
a tank can be measured remotely to 
+2 mm or better from outside the con- 
tainer by means of a fast-neutron source 
attached toaslow-neutron counter. The 
source may be positioned below but not 
too close to the counter, or above the 
counter but not too far away from it. 
Equal precision is attainable when the 
source is stationary and either opposite 
the liquid outside the tank or inside the 
liquid. 
* * * 
We are grateful to P. R. Malmberg for 
helpful advice, and to M. Giuliano and P. 
Szydlik for assistance with some of the 
measurements. 
BIBLIOGRAPHY 
1. D. G. C. Hare, U. S. Patent 2,323,128 (June 
29, 1943) 
2. L. S. McCaslin, Oil Gas J. 46, No. 21, 100 
(1947) 
. A. Wolf, U. S. Patent 2,456,233 (1948) 
. D. M. McCutcheon, Am. Foundryman 16, No. 
6, 35 (1949) 
. F. S. McCarthy, G. A. Rice, J. Electrochem. 
Soc. 97, 249 (1950). 
6. J. E. Johnston, Brit. Cast Iron Research Assoc. 
J. Research and Develop. 8, 523 (1950) 
7. R. Meakin, J. Sct. Instr. 28, 372 (1951) 
8. J. E. Jacobs, R. F. Wilson, Electronics 24, 
9 
a ww 
No. 10, 172 (1951) 
. D. G. C. Hare, U. S. Patent 2,348,810 (May 
16, 1944) 
10. J. H. Kunkel, Petroleum Engr. 16, No. 12, 155 
(1945) 
11. D. G. C. Hare, U. S. Patent 2,378,219 (June 
12, 1945) 
12. K. H. Sun, et. al., Phys. Rev. 95, 600 (1954) 
157 
