|Re: Response to Draft ICRP Document Lung Cancer Risk from Radon and Progeny
Dear Mr. Clement
Cameco is one of the world's largest uranium producers accounting for about 16% of the world's production. Cameco has controlling ownership of the world's largest high-grade uranium reserves and low-cost operations in northern Saskatchewan, Canada. We are also the largest US producer with in situ recovery operations in Wyoming and Nebraska. In Kazakhstan, production at the Inkai joint venture is ramping up to its initial approved capacity. Cameco holds premier land positions in the world's most promising areas for new uranium discoveries in Canada and Australia as part of an intensive global exploration program. Cameco is also a leading provider of processing services required to produce fuel for nuclear power plants, and generates 1,000 MW of clean electricity through a partnership in North America's largest nuclear generating station located in Ontario, Canada. Cameco is committed to having a safe, healthy and rewarding workplace and radiation protection is a key aspect of our efforts in this area. As such, we are keenly interested in ongoing developments in the understanding of the risks from radon progeny and how we can improve safety at our operations.
The draft ICRP report called Lung cancer risk from radon and progeny provides a comprehensive review of the ongoing epidemiology studies of radon, both at home and at work, notably in uranium mines. The draft report builds upon the latest UNSCEAR report on radon (UNSCEAR 2009) and comes to similar conclusions on the numerical risks of radon from combined residential studies and combined miner studies. Both these reports highlight an important development in our understanding of radon, that is, the pooled residential studies now provide a direct estimate of the risk from radon without the need to rely on dosimetric extrapolations from the miner studies. In this regard, a coherent picture is emerging on radon risks in two quite different environments, residential and industrial mining, and importantly the risks are based upon direct epidemiological evidence. While the numerical risk estimates will undoubtedly undergo further refinements as new and updated epidemiology studies are conducted, a compelling body of evidence now exists for policy makers on radon risks.
At a time when there is now direct epidemiological evidence of radon risks for both residential and mining environments, we believe adopting a dosimetric approach for radon to be premature and unnecessary. The draft ICRP document does not make a case for the efficacy of the dosimetric model for radon nor does it demonstrate a need for the use of the dosimetric model and finally does it not appear to have given sufficient consideration to the practical implications of adopting it. We believe that retaining the epidemiological based approach provides a strong basis for standard setting and urge the ICRP reconsider adopting the dosimetric approach for radon until further work is done in this area.
The ICRP reasons for adopting the dosimetric approach appears to rely on the desire for a coherent approach to other radionuclides and the apparent agreement between the dosimetric approach and the epidemiological one. With regard to the desire for a coherent approach to other radionuclides, we do not believe this is a compelling reason to change. For most other radionuclides, little direct epidemiological evidence exists of their cancer risks. However, there is a coherent picture of radon risks from direct epidemiological evidence and there is no need to use a dosimetric model for radon. Indeed, it is with some irony that when the dosimetric considerations needed in the past to quantify residential radon risks are no longer required, the ICRP proposes to abandon the epidemiological approach for the workplace, which has provided direct evidence of the radon risk for decades.
While the ICRP states that it wants to treat radon and radon progeny in the same way as other radionuclides, the reality is that some aspects of the models needed for radon progeny are not the same as for other inhaled radionuclides. The physics needed for the submicron radon progeny particles are not the same as those required for the much larger size particles typically encountered for other radionuclides. This means the models are different and more importantly the workplace characterization needed to accurately use these models are also different. Thus, in practical terms the adoption of a dosimetric approach for radon does not simplify the radiation protection system, but makes it considerably more complicated. At the current time it is difficult to justify adopting a significantly more complicated approach.
The draft ICRP report provides an extensive review of the epidemiology, but provides no evidence to support its proposed dosimetric model, other than the apparent overall agreement between it and the epidemiology studies. This in itself is not sufficient proof to demonstrate whether this is a coincidence based on a fortuitous choice of many parameters for the model or true agreement. The fact that the ICRP dosimetric model does not adequately account for smoking, which is by far the largest risk of lung cancer, casts a serious doubt on the model. When one considers that the range of risks from the epidemiology studies is about an order of magnitude, presumably some of this variation is due to the different aerosol characteristics in these different mines. A convincing proof of the efficacy of the radon dosimetric model, but admittedly difficult one, would be re-examine some of these studies with appropriate aerosol parameters and see if the range of calculated radon risks decreases. The fact that the model has not been calibrated and it is does not realistically account for smoking are indications that more work on it is needed prior to adopting it for decision making purposes.
The ICRP has stated it will provide dose coefficients for different reference domestic and workplace conditions. However, there is very little data available for modern uranium mines to provide such default conditions. The scarcity of data applicable to existing uranium mines in Appendix B points to this problem. There have been few measurements characterizing radon aerosols in uranium mines in the last 15 to 20 years. Over that same time there have been changes in mining in general and in the uranium mining industry in particular. For example, in Canada the large low-grade underground mines are now closed and a new generation of very high-grade mines are in operation. In addition, in situ recovery mines are now a very important source of uranium. Furthermore, there are few uranium mines left in Europe and the United States and new centres of production have risen. For example, Kazakhstan is now the largest supplier of uranium in world and most of its production is from in situ recovery mines. To our knowledge there is virtually no information from these important new uranium production centers that have been developed in the last 15 to 20 years. This means it is not possible to reliable establish default parameters based upon workplaces at the current time.
The lack of information on current uranium production facilities points to another problem. Currently industry, and we believe most regulatory agencies, do not have the ability to gather this data in a short period of time. There are several reasons for this. There has been little work on characterising radon aerosols over the last two decades. In addition, the complexity of these measurements makes it inherently more difficult to collect this data. Also, there is not a standard measurement protocol for radon aerosols that is widely accepted. In essence, the premature adoption of a dosimetric will place industry, and we believe most regulators, in a situation where they cannot effectively characterise the workplaces that have a responsibility for. This becomes a particularly acute problem because we do not believe there is sufficient information currently available to establish reliable default parameters.
We believe that for the time being there is no need to adopt a dosimetric approach for radon. The results from the latest epidemiology studies, as analysed by UNSCEAR and the draft ICRP report, provide a reasonable and stable basis for ensuring adequate protection of workers. Prior to adopting a dosimetric approach for radon more work is needed, both by the ICRP, to further validate and improve the model, and by industry, to characterise the nature of radon aerosol properties in our varied workplaces. With regards to the later endeavour a number of major uranium producers (i.e., AREVA, BHP, Cameco Corp., and Rio Tinto) are undertaking a project in this regard. Our first step is to establish, with the help of suitable experts, a standard measurement protocol and we would welcome the cooperation of the ICRP in this area. Following the establishment of the measurement protocol we plan to characterise our respective workplaces. Collectively our companies operate open pit mines, mills in hot and cold climates, low and high grade underground uranium mines, and insitu recovery uranium operations. We estimate that this work will take several years to complete. We strongly believe that the input from this work is necessary prior to adopting a dosimetric model for radon and publishing default dose conversion factors.
We support the ICRP’s ongoing efforts to improve the system of radiological protection and believe the ICRP draft report Lung cancer risk from radon and progeny illustrates the progress that has been made in understanding radon risks with epidemiology. However, it is premature to abandon this approach when much work still needs to be done on the dosimetric approach to radon. We look forward to working with the ICRP and other interested groups in this area.
Director, SHEQ Systems