R G Lyons1, S B Solomon, R Langroo, J R Peggie2
1 Department of Physics, University of Auckland, Private Bag 92019, Auckland, New Zealand
2 Australian Radiation Laboratory, Lower Plenty Rd, Tallambie, Victoria 3085, Australia

Last year saw the completion of our major study of radon in Australian tourist caves, copies of which have been forwarded to all participating cave management bodies (Solomon et al. 1996).  This year has seen the completion of the corresponding study in New Zealand caves.  Reports summarising the data are currently in preparation and each cave management that participated will receive the information relevant to their caves when these are completed.  The paper on the Australian study presented at the last ACKMA conference in Tasmania (Lyons et al. 1997) included background information which is equally relevant to this New Zealand study.  As some of you may not have ready access to the Proceedings of the Tasmanian conference, a summary of the background information is included here, but if you have a sense of deja vu, skip to the section on Methodology.

First, let's be clear about what precisely the risk of exposure to radon may be.  The risk is NOT that anyone exposed to radon will develop acute radiation poisoning at the time of exposure.  The risk is that, with prolonged exposure to high levels, a person will have an increased risk of developing lung cancer or other cancers of the respiratory track in the future.  In this, it is similar to smoking — it doesn't follow that in either case, the person WILL develop cancer, or that, if they do, the cancer will be due to exposure to cigarette smoke or radon, simply that exposure to either substance will increase the risk of developing cancer.

The relationship between low levels of exposure to radon or any other radioactive substances, or to environmental factors such as Agent Orange or smoking, is always difficult to establish precisely.  It has taken years of extensive studies involving large numbers of people to convince people of the link between smoking and lung cancer, indeed the tobacco companies still claim that the evidence is "inconclusive".  This difficulty of proof is inherent in any study of environmental factors, and in the case of radon in caves the numbers of people involved are too low for statistical tests to be useful.

Because we can't prove radon is a cause in any particular case of lung cancer, does that mean there is no risk?  Our concerns are the result of the study of radiation, such as X-rays, nuclear fall-out and medical applications of radiation, where health risks have been clearly demonstrated for higher doses.  From these proven links for high dose cases, scientists extrapolate to determine the much lower risk for much lower doses.

Regulations

The level which the International Commission on Radiological Protection (ICRP) proposed as a "safe" level, below which no action is necessary, corresponds to 2-3 times a normal "background" level, to which the average person is exposed in normal daily living. These recommendations are in the process of being adopted by individual national governments such as Australia and New Zealand.  Specifically they are that, when a working environment has concentrations of radon which exceed 1000 Becquerels per cubic metre (Bq m-3), either levels must be reduced by intervention such as ventilation, or people working in that environment must be monitored to ensure the total dose they receive at work is less than 20 mSv per year.  Comparable regulations formulated by the National Radiation Protection Board (IRPB) in the United Kingdom are expressed in a different radiation unit called the Working Level (WL).  Converted into Bq m-3, they specify (approximately) 100 Bq m-3 for ionizing regulations to be applied, 200 Bq m-3 for Government action level for radon in houses and 400 Bq m-3 for a Controlled Area to be designated. 

At the maximum levels of exposure recommended by the ICRP, the increased risk is comparable to that of smoking 2-3 cigarettes a day, or to passive smoking.  Why then, should we be concerned?  Particularly when many cave workers are heavy smokers and indulge in other hazardous activities, such as driving on our roads or even caving?  The difference is clear-cut.  If people are exposed to increased risk of any sort through their work, this is ethically and legally very different from a voluntarily assumed risk — the Marlboro Man is currently suing his erstwhile employer because he has developed lung cancer which he maintains may be due to his smoking cigarettes in making the advertisements. 

Measurement of radiation levels, and radon in particular, are not straightforward.  Because the health effect of radiation depends on many things, not just the crude measurement of how much radiation there is, but also on how long the person is exposed to the radiation, what type of radiation it is, where it is absorbed in the body (in the case of radon and its products, in the lungs themselves), and on other external factors such as the health, age and lifestyle of the person, radiation dosimetry is very complex.  It's important to understand one thing in particular, though: the risk depends on the TOTAL dose, which is the concentration multiplied by the time during which the person is exposed — a high concentration for a short time has exactly the same risk as half the concentration for twice the time.  The guidelines are based firstly on a simple easily carried out measurement of overall levels, below which it is considered that there is no cause for concern and no more complex monitoring needs to be carried out.  At these levels even if a person spent 2000 hours per year working in the area their total dose would still be less than recommended maximum for occupational exposure levels for licensed nuclear radiation workers.  If these levels are exceeded, then personal monitoring needs to be undertaken which will give the total dose received by that individual during the time they actually spend underground.  If these also show high risk levels, then further action needs to be taken (see later).  Personal monitoring, though more informative, is more expensive than simple measurements of average radon concentrations; where the simpler measurements of radon levels are low enough, clearly it's easier for all concerned.

Motivation for the New Zealand study

As for the Australian study, the motivation behind this study was NOT to disrupt the operation of tourist caves, nor to provide scientists with things to do — we, like you, have plenty to keep us busy! — but to avoid blind bureaucratic sledge hammers and public panic. The alternative, of course, was to wait for the regulations to be passed — and I remind you that the recommended regulations were initiated by the ICRP quite independently — sit back and collect consultancy fees!   Given that it is a legal responsibility of employers to provide a safe environment for employees, and that some overseas caves had been shown to have elevated levels of radon, it would be rash for any operation not to carry out monitoring and this additional expense might well have strained the resources of smaller operations.  The advantages of a single Australasian wide study were:
? Economies of scale, both financial and in effort.
? Directly comparable results from all caves measured, due to a common method and common calibration standard. 
? The possibility of external funding, relieving smaller and remote operations, in particular, of a disproportionately heavy expense.
? Further work, if necessary, could be targeted to those areas which had the potential to result in high doses to employees.
? The provision of data on which wise management decisions could be taken.  Caves vary so widely that data obtained from one cave simply may not be applicable to others.
? The study would be the most comprehensive and geographically diverse study of radon concentrations in caves to date and would add significantly to our knowledge of radon in caves on an international basis.

In the event, funding was obtained for the Australian part of the study from the Australian Occupational Health and Safety Group, Worksafe.  The New Zealand study has piggy-backed on the Australian study, using the same technology with monitors provided and processed by the Australian Radiation Laboratory.  Funding for the New Zealand study has come from a personal research grant from the University of Auckland and field-work support in cash and kind by cave owners and management. 

Methodology

Details of the methodology are given in the full report of the Australian Worksafe study (Solomon et al. 1996).  In summary:
• In New Zealand, a total of 112 sites in 22 cave systems were monitored. 
• The monitors were passive track etch detectors, calibrated by Australian Radiation Laboratory.  Four monitors in each site covered the four seasons, while an additional monitor remained in situ for the whole year to collect an annual average and provide a cross-check with the seasonal data.  Sites were selected to be representative of each cave, and the times which tour guides spend in the areas of the caves represented by each site were also noted.
• Track etch monitors record average radon concentrations.   They are relatively cheap and it was thus possible to deploy the large numbers necessary for a comprehensive study such as this.  However, the potential health risk is due, not to radon itself, but to radon progeny, which are produced by radon when it decays radioactively.  Radon concentrations give an upper limit to the concentrations of radon progeny which may be present: sites which are low in radon must also be low in radon progeny, so we can give such sites a "clean bill of health".  If the average radon values are high, these caves will require further work, and this is where future efforts should be targeted. 
• Track etch monitors record average radon concentrations over the time for which the monitors are exposed, in this case 3 month seasons. However, average seasonal values may not be typical of the times at which most tours occur, e.g. day and night values may be very different, but this will not be observed in seasonal averages.
• Radon progeny measurements were also taken on two separate occasions in each site using a Thompson and Neilsen Instant Radon Progeny Meter.  This assisted with identifying areas of the caves in which the air had similar characteristics (and therefore could be represented by a single monitor) and those which were significantly different, particularly areas of high concentrations.  The radon progeny measurements will also assist with estimating actual doses and the likely health risks.

Results

The average seasonal concentration of radon varies widely, ranging from background levels to more than 21,000 Bq m-3 (Figure 1)
• Approximately 40% of the values are less than 400 Bq m-3, which is comfortably below the levels recommended by the ICRP for intervention and equal to the statutory limits for a designation of a radiation control area set by the National Radiation Protection Board (IRPB) in the United Kingdom. The lowest values recorded are several times typical outdoor background values but at less than 100 Bq m-3 are negligible; being below the level at which IRPB regulations apply — someone could not only work full time in these areas without exceeding the yearly recommended maximum dose, they could also live there.
• A further 24% lie between 400 and 1000 Bq m-3  As these fall below the range at which action is required according to the ICRP, this is generally reassuring.  However, they are above the level at which the IRPB regulations for the United Kingdom require a Radiation Controlled Area to be designated and monitoring to be carried out.  In some cases these levels are of particular concern, because data for the same sites in other seasons is lacking.  Given that the values we know are elevated and knowing also that seasonal variation can be very great, it is quite possible that even higher values occur for the undocumented seasons. 
• The remaining 36% of sites have average seasonal concentrations greater than 1000 Bq m-3 which is the level at which the proposed regulations require action to be taken, either to reduce the concentrations or to monitor workers to ensure recommended radiation doses are not exceeded.  The highest value recorded was 21,000 Bq m-3, more than 20 times the recommended maximum.  However, as time spent at a site is also a factor in the radiation dose received, a concentration as high as this will not necessarily result in an excessive exposure to radiation.
• Spatial variation. All areas showed some sites with high concentrations.  In many caves, high values in some sites were balanced by lower values in others, giving an overall average below 1000 Bq m-3 , but in some caves most of the sites were high, yielding unacceptably high averages over the cave as a whole for some times of the year.
• Seasonal variation.  Figure 2 gives the average radon concentrations by season.  Overall, average values are lowest in autumn, and highest in spring.  Spring and winter have the most sites with concentrations above 1000 Bq m-3 , with 49% and 44%, respectively.  However if all sites above 400 Bq m-3  are included, spring and summer have the highest values, with 77% and 73% of sites having elevated levels.  These general statements should NOT be taken as a guide for management: there are very significant exceptions to this general trend in particular caves, which may be explained in terms of the different factors governing the air circulation in caves of different configurations.  Generally speaking, summer and winter are the more stable periods with respect to air circulation within caves, which is shown by more of the values falling into the highest and lowest classes and less in the intermediate 400 - 1000 Bq m-3 class.  Autumn and spring reveal the more volatile nature of the cave air dynamics with a graduation of values from lowest (most frequent) to highest in autumn or highest (most frequent) to lowest in spring.
• In any one cave there may be very marked seasonal variation, by a factor of more than 50 — low concentrations in spring, for example, do not mean that there will be low concentrations in winter, nor do high values in one season imply high values in another.  Because of this, it is imperative that data be collected for the full year. It is unfortunate that some caves missed out on this.  The diligence of many cave managers in collecting and forwarding the monitors for analysis is rewarded by the quality and usefulness of the data obtained for their caves.
• The comparison of annual data with seasonal data shows that the track etch detectors may significantly underestimate radon concentrations for longer periods, as annual values are sometimes substantially lower than the sum of the 4 contributing 3 month periods.  It is thus unfortunately not possible to estimate missing seasons by subtracting the values for the measured seasons from the annual total.  It also suggests that the 3 month averages may be underestimated as the 3 month exposure time is longer than the calibration exposure time.  Australian Radiation Laboratory is currently carrying out further calibration trials which should establish the extent of any underestimation and enable the data to be revised upwards if necessary.

Figure 1: 
Frequency histogram for 3 monthly averaged concentrations of radon for sites in New Zealand tourist caves.

Figure 2:  Seasonal distributions of average radon concentrations for sites in New Zealand tourist caves

Figure 3:  Frequency distributions of 3 monthly averaged radon concentrations in (a) New Zealand, and (b) Australian tourist caves.

Comparison with Australian Results

Overall, the New Zealand sites measured have somewhat higher concentrations of radon than do the Australian sites (Figure 3), with 36% exceeding the 1000 Bq m-3 level compared to 28%.  If sites above 400 Bq m-3 are included, 59% of the New Zealand cave sites measured have these elevated levels compared to 53% of the Australian cave sites.  The highest value recorded in New Zealand, 21,000 Bq m-3 was more than twice the highest Australian value. 
There are a number of possible contributing factors: the volcanic sediments and other uraniferous rocks associated with many New Zealand caves, which provide a good source of radon; the grainsize of sedimentary deposits within the caves and their moisture content, which affect the rate of diffusion from the parent matrix into the cave air and, thirdly, the degree of ventilation of the caves. These factors are discussed more fully in a paper to be presented to the South Pacific Radiation Association in February, 1998.

Implications and Possible Actions

For caves which have annual averages (over all sites) monitored above the maximum recommended of 1000 Bq m-3, there are a number of possible management options.  The best line of approach will depend on the particular cave and its usage pattern. 

• We can close the caves and all go home — a knee-jerk response and quite unnecessary.
• We can ventilate the caves to dilute the radon. In a gathering such as this I don't need to spell out the possible adverse effects of this on the cave biota and the speleothems which are one of the principal attractions of many of the caves. NOT RECOMMENDED!

Not only is increased ventilation undesirable for microclimatic reasons, but it is also not always effective and may in fact even draw more radon into the cave air.  This has happened in a case in Margaret River in Australia, when the engineers did not realise that the cave’s natural ventilation reversed seasonally.  Thus for part of the year the forced ventilation supplemented the natural ventilation, reducing radon levels as desired. However, when the cave's natural ventilation reversed with the seasonal variation in temperature, the forced ventilation acted against the natural ventilation and radon concentrations were increased above the values they would have had without any attempt to force ventilate.  Any attempt to alter or increase the ventilation in a cave MUST be preceded by a thorough understanding of the principles governing that particular cave's air circulation system or the last state may be worse than the first.

• The probable dose received by an employee can be calculated and action taken if this, rather than the concentrations, are high.  When the actual hours worked are taken into account, the dose due to exposure even to relatively high levels of radon, may still be low enough not to be of concern.
• Personal dosimeters can be provided and employees taken off underground duties when their accumulated dose reaches the maximum permitted — a reactive management response.  Effective in fulfilling legal requirements and ensuring employee safety standards are met, but may have undesirable social consequences, or make rostering of duties difficult. 
• Estimates can be refined so they are more accurate.  For example, standard factors used to convert radon concentrations into dose from radon progeny can be actually measured instead of assuming a value based on measurements taken in houses and mines.  This work is currently in progress.
• Where tourist operations are large and there are a number of cave tours and employee duties, there may be a number of management options which will reduce the radiation dose to any one employee.  Such options might include choosing tours which are expected to have a low dose in a particular season, rostering staff with a high number of hours underground to lower-dose tours, and scheduling maintenance work for low radon seasons.  To do this effectively requires more in-depth research to find out how, when and why radon and radon progeny concentrations vary, in greater detail than the relatively coarse estimates provided by average seasonal radon data.  The more we know and understand, the greater the opportunity for proactive management.  And the less we will be restricted by the necessary safety regulations. 

More than a third of sites measured were above the 1000 Bq m-3 level at which the International Commission for Radiological Protection recommend further action to monitor and reduce radiation exposure should be taken.  All areas in New Zealand have some sites with high values.  If a new cave in any area is to be opened up for tourism it cannot be assumed that it has negligible values and it should be checked.  However, the data show that for many New Zealand caves average concentrations are sufficiently low that no further action needs to be taken.  Other caves will require further work, depending on the usage and the employee hours worked in the caves.  For some caves where the data are incomplete, additional base measurements such as those carried out in this study must be carried out to obtain enough data to assess the risk and determine an appropriate course of action.
The options available and the optimal response will depend on the particular cave, the levels of radon and the patterns of employment — it would be rash to generalise.   Nevertheless, the information gained in this study, both the radon data and the more comprehensive fieldwork investigations, is an excellent basis for discussion and development of appropriate measures for specific caves where necessary. 

Acknowledgements

A study such as this is simply not possible without the co-operation of those on site.  Special thanks are due to the many of you who have contributed in so many ways, with changing the seasonal monitors, with hospitality and practical help. I am also most grateful for feedback and the careful observations many of you have contributed.  Your knowledge of your own caves is extremely valuable, and not something that can be easily obtained in a short visit to your caves.

References

Lyons, R G, Solomon, S B, Langroo, R, Peggie, J R & James, J M  1996, Occupational exposure to radon in Australian tourist caves,  ARL/TR119 

Lyons, R G, Solomon, S B, Langroo, R, Peggie, J R & James, J M  1997,  Radon monitoring in Australian tourist caves: the why what and wherefore of the Worksafe study, Cave and Karst Management in Australasia 11: Proceedings of the eleventh Australasian Conference on Cave and Karst Management, ACKMA and NSW Parks and Wildlife,  Hobart  pp.197-201