Environmental Geochemistry and Health 1993 15(2/3) page 145

Radon in soil gas and its relationship with some

major faults of SW England N.R, Varley* and A,G. Flowers School of Applied Physics, Kingston University, Penrhyn Road, Kingston-Upon-Thames, Surrey KT1 2EE, England

Abstract

The south-west of England was designated by the National Radiological Protection Board (NRPB) as the first 'Radon Affected Area', as over 1% of the housing stock is estimated to have an indoor radon concentration in excess of the 200 Bq m-3 Action Level. The situation is even worse for houses situated above uraniferous granite intrusions, where over 30% are thought to be above the Action Level. The aim of this study is to investigate the relationship between the level of radon in soil gas and the local geology. Particularly high radon levels were measured along major fault zones. This could be explained by: increased rate of migration of the radon due to the permeable fault, the presence of radium or radon-bearing ground water within the fault, or secondary uranium mineralisation. Seasonal variations are also considered.

Introduction The radiological hazard associated with radon gas was first identified in mine workers. In the seventeenth century, it was first suggested that a gas breathed by miners often led to a disease of the lungs. However, it was not until the early 1970s that the potential hazard from the inhalation of radon gas in the domestic environment was first identified. Since then, many houses in the UK have been monitored in several surveys conducted by the NRPB and the local authorities. At the end of 1991, 92,000 dwellings had been measured for indoor radon: 12,000 houses with concentrations above the Action Level had been identified out of an estimated 100,000 in England (Green et al., 1992). There are two isotopes of radon that are significant contributors to the radiation dose received by the population 222Rn, which is part of the 238U decay scheme, is the most important as it has the longer half-life of 3.82 days. 22°Rn is a decay product of232Th and has a half-life of 55.6 seconds. Both decay sequentially to several further alpha emitters, which can then be deposited in the lungs, resulting in damage to the thin internal membranes. Radon levels in SW England The latest estimates by the NRPB are that the average annual radiation dose in the UK from radon and its daughter products is 1.2 mSv, which represents 47% of their total annual dose. Whereas in Cornwall the figure is 9 mSv or 87% (Green et al., 1992). To whom all correspondence should be addressed.

The area with the highest proportion of homes above the Action Level of 200 Bq m-3 is the south-west of England. This is also the area in which most of these homes have been identified (by the end of 1991 the number was 11,000, or 18%, of the estimated 60,000 present in the two counties) (Green etal., 1992). H o w e v e r , many more measurements will be required before the full extent of the problem can be established, in accordance with the government's goal to rectify the radon problem by the year 2000 (NRPB, 1992). In 1990, the NRPB designated the south-west of England as the first 'Radon Affected Area' (NRPB, 1990). This is an area in which over 1% of the housing stock is estimated to have an indoor radon concentration in excess of the 200 Bq m-3 Action Level. The distribution of houses with a high indoor radon concentration is largely controlled by geology. The 'potential' for elevated indoor levels is also influenced by characteristics of the soil, particularly its permeability. However, it is the quality and type of house construction that finally controls the indoor concentration of radon in a particular house. This represents the main problem when attempting to predict indoor radon levels. The other is obtaining the appropriate data. Unless the home is situated above a fault, the scale of a geological map will usually be too small to show sufficient detail of localised variations either in uranium content or in other factors that may affect the permeability (e.g. the degree of fracturing). Even if a c c u r a t e data on the u r a n i u m concentration and permeability of the underlying rock and soil have been obtained, it has often proved

146

Radon in soil gas

Granite plutor~

D

Dartmoor

B

Bodmln Moor

A

St. Am, tell

C

Canmaendlb

Possible underground granite Fault zone

"r" - ' r "r"

L

land'. End

S

Sally Ide,

II

Hingston Down

K

Kit I-rail

Lizard -

:

- Start fault

Dodman

t

X

......-"

Sfir~]e.l~lh \ ~,s.~ z~

--

/

---

Fault

Position of traverse~ ...... Limit of metamo~hle aureole -,....

~ i

It

Sdcklepath Fattlt

~

\

I

J

0 I

I

0

10

20

\

~

30

20

iO

J

I I

I

30

40

I

N

i

40 M i l e s

I

I

I

5o 6o

Kilometres

Figure 1 Location of granite in South-West England.

impossible to predict with any great precision the levels of radon indoors. Surveys often show unexpected results with poor correlations between indoor radon concentrations and the amount of uranium in the rock beneath (Sachs et al., 1982; Fakundiny and Friedman, 1988; Voutilainen et al., 1988; Damkjaer and Korsbech, 1988). It should however be possible to predict the maximum likely indoor radon concentration.

?--

Geology of SW England The south-west of England is dominated by a granite batholith. The granite is uraniferous, with a concentration of 15-20 ppm. (Ball et al., 1982), and often highly fractured or weathered. It has intruded through the Carboniferous and D e v o n i a n sedimentary rocks to be exposed in five major outcrops (Figure 1 modified from Edmonds et al., 1975). Over 30% of homes located above the granite have been found to have an indoor radon concentration above the Action Level.

3CCCOC Reported position

E

of [auJt

I

25CCOC c~

Granite

2cocce b 'S ]SCCCO

?

'~

Io~c'o

° 500 0

/ 0

"\." /

b-4,

200

,'

~CC

\& 0 6C0

800

1OOC

0is:¢~ce (m)

Figure 2 Traverse across Sticklepath Fault 1.

12C0

14CC

Uranium exists in the granite as an accessory mineral, but in some locations it has been concentrated into vein-type mineralised deposits. Re-mobilisation of the uranium can also occur along faults or fractures. This has occurred along some of the wrench fault zones that cross much of the south-west peninsula of England (Blyth, 1957;

N.R. Varley and A.G. Flowers

200000

P o s i [ i o n of main fault

180000

I

147

,

I

...........

160000 m

500

I 22z

•

Rn

,

Rn 22°

400

140000 7a

0

120000

P o s i t i o n of branch fault

300

100000 o

0 o

80000

/ --

~

I'iQ 0 "0

200

tt t

60000

I

40000

100

q~/

20000 0

0

100

It""

"r"

i~,,"'m"',,,im;

t

I

I

I

r

200

300

400

500

600

lit

0

Distance (m) Figure 3 Traverse across Sticklepath Fault II.

Dearman, 1963; Turner, 1984). Early studies showed that the radon level in soil gas can increase above a fault (Israel and Bjornsson, 1967). Also, in some surveys carried out in the US, higher indoor radon concentrations have been recorded in homes situated above faults (Ogden et al., 1987; Gundersen et al., 1988; Steck, 1988). As well as a possible increase in the amount of radium decaying to produce the radon, a fault can offer radon a convenient route in its migration to the surface. Soil gas m e a s u r e m e n t s

Large variations in the radon concentration of soil gas can occur, both in time and over small distances. The former is due to meteorological effects, the most influential being rainfall, as the level of moisture in the soil greatly affects the radon concentration in a number of ways (Rose et al., 1990a). The main effect is a reduction in the rate of diffusion of the soil gases through the pores, as the soil pores become filled with water. A layer of water near the surface of the soil can also have a capping effect, greatly reducing the emanation of radon at the surface. The variation in soil gas radon concentration over small distances will usually be due to the inhomogeneous nature of soil. Its permeability can vary greatly over only a short distance. Finally, there is the variation in concentration with depth. This does not follow a simple diffusion model and some complex variations have been observed (Kristiansson and Malmqvist, 1984; Rose e t a l . , 1990b; Varley and Flowers, 1991). The concentration of radon in soil gas is less affected by

meteorological conditions at greater depths; hence it would be ideal to sample at depths of a metre or greater. However, within a given time it may be more beneficial to take a greater number of shallow measurements, than a few deep measurements. In any case, it will often be too difficult to penetrate to any great depth.

Methodology

Levels of both gZgRn and 22°Rn were measured using Lucas cells and a Pylon AB-5. The two isotopes were differentiated using a method developed by Morse (1976), in which counts due to 2g°Rn and its d a u g h t e r s are s e p a r a t e d by consideration of their much shorter half-life. The Lucas cells were calibrated using facilities at the National Radiological Protection Board, where the cells were exposed to a known concentration of radon. The soil gas was pumped to the surface over a one minute period through a probe with an internal diameter of 10 mm. The probe was usually placed in the soil to a depth of 50-60 cm, unless the depth of overburden was less. Track etch detectors were used to take longer integrated measurements of the radon concentration in the soil. They were buried to a depth of 30-60 cm and left for between one and two weeks. Traverses were taken across fault zones with a spacing of between 20 and 50 m. In places a shorter spacing was used to increase the definition of any peaks. Several traverses were repeated at different times of the year and in different meteorological conditions.

148

Radon in soil gas 800000

,

,

,

position

Reported

t

J

28/.3/91

600000 i

E C3co

400000

-

6

200000

-

/

c © o

O~

Oq C'q oq

1 c

w

600000

-

400000

-

200000

-

Depth

IX.

of_ co CD

•

,/

• •

52 cm } .30 c m

17/7/91

•

52 cm

.30/7/91

7--. /.--%•

.__-,

-%

0 O9

_

200000

-1

I

E

150000

-

100000

-

50000

-

Uco

r4 LJ

0 0

I

I

I

50

100

150

Distance

I

200

(m)

Figure 4 Traverse across Sticklepath Fault IlL

Results The concentration of 222Rn in the soil gas varies greatly across the south-west of England. The highest concentrations were measured in areas dominated by granite intrusions, with a mean level of about 150,000 Bq m -3. The maximum level measured was over 900,000 Bq m -3, which was above the Sticklepath Fault (see Figure 1) in an area where it passes through granite. Figure 2 s h o w s the results of the 222Rn concentration in the soil gas along one traverse across the Sticklepath Fault. The d e c r e a s e d

permeability of an alluvial deposit in the valley had the effect of reducing the radon concentration measured in the soil gas to zero. This accentuated the effect of the fault. In certain locations, smaller faults branch off from the S t i c k l e p a t h Fault. A t r a v e r s e was performed in one such location, the results of which are shown in Figure 3. As well as 222Rn, results of the measured concentration of 22°Rn are shown. The two isotopes largely showed a similar pattern, with a greater enhancement of the 222Rn concentration over the main fault. Where measurements were repeated at the

N.R. Varley and A.G. Flowers

same location on more than one day, a variation in concentration was observed. This is shown at one location in Figure 4. During the winter, when the level of moisture in the soil was greater, no radon was measurable in several positions along the

500000

i

i

I

i

149

traverse. However, above the fault higher concentrations of 222Rn were recorded at this time of the year. The effect of sampling depth is also shown. Similar trends were observed from measurements made at different depths. I

I

e~

1

I

I

I

I

I

I

400000

29/5/91 300000

200000

100000

0

500000 I

400000 d o o

I

15/8/91

300000 200000 100000 V .... ~ "'Y . . . . ~r'"

0

Q

50O000 [

400000

• 55cm depth

ooooo °°°°°1ooooo .

.

:

0

I

0

200

400

600 DisLance

Figure 5 Traverse across Prewley Fault (South Down).

800 (m)

1000

1200

1 400

150

Radon in soil gas

Along this traverse track etch detectors were used to measure the radon concentration over a period of just under two weeks. Grab- samples were taken at the beginning and end of this period at each point. A peak at 150 m from the beginning of the t r a v e r s e was o b s e r v e d using both methods. However, the peak at about 55 m was not revealed by the track etch detectors. The Prewley Fault also crosses the northern boundary of the Dartmoor granite mass, as shown in Figure 1. In Figure 5 it can be seen that much higher radon concentrations were recorded on the first and third repetitions of the traverse. The two days on which these measurements were made both followed a period of rainfall. The results from the track etch detectors were far less conclusive, but a peak was still observed in the same position as recorded from the grab-sampling results. Discussion

Many closely spaced traverses were taken across the Sticklepath and other faults. Consistently, higher levels of 222Rn were measured above the faults. It has been suggested that the Sticklepath Fault represents a zone of descending hydrothermal fluids and as such has low levels of radon associated with it (Durrance and Heath, 1985). Unlike 222Rn, enhanced concentrations of :2°Rn were not always found to occur over a fault. The half-life of 22°Rn at 55.6 s is much less than that of 222Rn. Therefore this isotope will not be able to migrate such great distances from its source. Consequently, the levels measured in the soil gas are lower and far more susceptible to changes in ~2ermeability. The absence of enhanced levels of °Rn, at the location of a peak in concentration of 2g2Rn, would indicate low permeability of the local soil. The measurement of radon in soil gas proved to jbe useful in determining the position of a fault. In several locations the technique enabled the position of the fault to be m o r e a c c u r a t e l y determined than that reported on existing maps. In Figure 5 the peak in radon concentration may suggest the inaccuracy to be as much as 500 m, h o w e v e r the larger peak could be due to an unrecognised fault. Generally, higher concentrations were measured in the soil over the fault zone during the winter months, but in surrounding areas lower levels were often recorded. An increased peak compared with a lower background means that improved results are likely if a survey is carried out during the winter. In previous studies of radon in soil gas, in other parts of the world, contrasting results have been obtained. Some have reported higher concentrations of radon at shallow depths during the winter (Rose et aL, 1990b; Asher-Bolinder et al., 1990), whilst others have recorded lower values (Sextro et al., 1989). Seasonal variations in the radon concentration in soil gas are mainly due to changes in the moisture

content of the soil. During the winter increased moisture levels lead to a higher proportion of the radon becoming dissolved. Lower temperatures also increase the proportion of radon becoming dissolved in the water within the soil pores (Rose et al., 1990b). For high moisture levels the soils can become water-logged, with all the pore spaces filled with water. No radon can then be measured. Within a fault zone, a greater volume of water will be present during a wet season. Where the water passes through the granite, it will probably have an increased radium content. This ground water is likely to be one reason for enhanced concentrations of radon over a fault. This was most graphically illustrated with the P r e w l e y F a u l t t r a v e r s e (Figure 5), where the peaks in 222Rn concentration over the possible fault position increased by a factor of two following rainfall. Due to the variation of radon in soil gas with depth, it would be preferable to measure the actual emanation rate of the radon from the ground. However this type of measurement is much more difficult to achieve, and the same large number of m e a s u r e m e n t s taken each day would not be possible. Lower concentrations of radon were measured by t r a c k e t c h d e t e c t o r s c o m p a r e d to the grab-sampling method. This could in part be due to the decay of the radon, during its migration to the d e t e c t o r cavity. This has been shown to be significant (Tanner 1991). Conclusion

Enhanced levels of radon were measured in the soil gas above some of the major faults of south-west England. The highest levels above the faults were measured during the winter or after rainfall. This is likely to be caused by a greater volume of ground water within the fault zones, containing relatively high concentrations of radium. Measurement of radon in soil gas using passive track etch detectors produced lower concentrations compared to the grab- sampling method. References

Asher-Bolinder, S., Owen, D.E. and Schumann, R.R. 1990. Pedological and climatic controls on 222Rn concentrations in soil gas, Denver, Colorado. Geophysical Res. Lett., 17(6), 825-828. Ball, T.K., Basham, I.R. and Michie, U.McL. 1982. Uraniferous vein occurrences of South-West England - paragenesis and genesis. Vein-type and similar uranium deposits in rocks younger than proterozoic, IAEA-TC-295/9, pp.113-158. Blyth, F.G.H. 1957 The Lustleigh fault in north-east Dartmoor. Geological Magazine, 94, 291-296. Damkjaer, A. and Korsb~h, U. 1988. A search for correlation between local geology and indoor radon concentration. Radiat. Protect. Dosim., 24, 51-54. Dearman, W.R. 1963. Wrench-faulting in Cornwall

N.R. Varley and A.G. Flowers and South Devon. Proc. Geological Assoc., 74(3), 265-287. Durrance, E.M. and Heath, M.J. 1985. Thermal groundwater movement and radionuclide transport in SW England. Mineralogical Magazine, 49, 289-299. Edmonds, E.A., McKeown, M.C. and Williams, M. 1975. South-West England. British Regional Geology. HMSO, London. Fakundiny, R.H. and Friedman, G.M. 1988. Workshop on geology and radon. North-eastern Environmental Science, 7, 63-69. Green, B.M.R., Lomas, P.R. and O'Riordan, M.C. 1992. Radon in dwellings in England. NRPB-R254. HMSO, London. Gundersen, L.C.S., Reimer, G.M. and Agard, S.S. 1988. Correlation between geology, radon in soil gas, and indoor radon in the Reading Prong. In: Marikos, M.A. and Hansman, R.H. (eds), Geologic Causes of Natural Radionuclide Anomalies, pp.91-102. Georad Conference. Israel, H. and Bjornsson, S. 1967. Radon and thoron in soil air over faults. Z. Geophysics, 33, 48-64. Kristiansson, K. and Malmqvist, L. 1984. The depth-dependence of the concentration of radon-222 in soil gas near the surface and its implications for exploration. Geoexploration, 22, 1741. Morse, R.H. 1976. Radon counters in uranium exploration. Proc. Exploration for Uranium Ore Deposits, Vienna, pp.229-239. NRPB (National Radiological Protection Board). 1990. Board statement on radon in homes. Doc. NRPB, Vol. 1, No. 1. NRPB (National Radiological Protection Board). 1992. Radon 2000 meeting. London. Ogden, A.E., Welling, W.B., Funderburg, R.D. and Boschult, L.C. 1987. A preliminary assessment of factors affecting radon levels in Idaho. In: Graves, B. (ed.), Radon, Radium and Other Radioactivity in Ground Water, pp.83-96. NWWA, New Jersey, USA.

151

Rose, A.W., Ciolkosz, E.J. and Washington, J.W. 1990a. Effects of regional and seasonal variations in soil moisture and temperature on soil gas radon. 1990 International Symposium on Radon and Radon Reduction Technology, p.3. EPA Paper C-VI-5. Rose, A.W., Hutter, A.R. and Washington, J.W. 1990b. Sampling variability of radon in soil gas. J. Geochemical Exploration, 38, 173-191. Sachs, H.M., Hernandez, T.L. and Ring, J.W. 1982. Regional geology and radon variability in buildings. Environment International, 8, 97-103. Sextro, R.G., Nazaroff, W.W. and Turk, B.H. 1989. Spatial and temporal variation in factors governing the radon source potential of soil. 1988 Symposium on Radon and Radon Reduction Technology, p. 1, EPA. Steck, D.J. 1988. Geological variation of radon sources and indoor radon along the south-western edge of the Canadian Shield. In: Marikos, M.A. and Hansman, R.H. (eds), Geologic Causes of Natural Radionuclide Anomalies, pp.17-23. Georad Conference. Tanner, A.B. 1991. Error in measuring radon in soil gas by means of passive detectors. Nuclear Geophysics, 5(1), 25-30. Turner, P.J. 1984. Hercynian high-angle fault zones between Dartmoor and Bodmin Moor. Proc. Ussher Society, 6, 60-67. Varley, N.R. and Flowers, A.G. 1991. Radon and its correlation with some geological features of the South-West of England. Natural Radiation Environment V, Salzburg. Voutilainen, A., Castren, O., Makelainen, I., Winqvist, K. and Arvela, H. 1988. Radiological characteristics of a village on uraniferous granitic ground in Finland. Radiat. Protect. Dosim., 24, 333-337. (Manuscript no. 295: received August 17, 1992; accepted after revision February 1, 1993)