Research report summaries 2004–2005

This Web page has been archived on the Web.

Archived Content

Information identified as archived on the Web is for reference, research or recordkeeping purposes. It has not been altered or updated after the date of archiving. Web pages that are archived on the Web are not subject to the Government of Canada Web Standards. As per the Communications Policy of the Government of Canada, you can request alternate formats on the Contact Us page.

Contractors’ reports are only available in the language in which they are submitted to CNSC.

RSP-0177 – Data collection for the ICDE (International Common-Cause Data Exchange) Project

This report describes the activities and methods used to collect data on switching devices and level transmitters in support of Canada’s data submission to the international Common Cause Data Exchange (ICDE) project. The ICDE project was established by the Organization for Economic Cooperation and Development (OECD) to encourage multilateral cooperation in the collection and analysis of data relating to common cause failure (CCF) events at nuclear power stations. The CNSC established its own project to collect data on Canadian nuclear power plants and fulfill its obligations under the ICDE agreement. The process used for collecting data on switching devices and level transmitters from Canadian nuclear power plants is included. The methods employed to analyse the data and code it into the ICDE database are described and summary statistics on the data are provided.

Top of page

RSP-0178 – Feasibility study: Saskatchewan uranium miners cohort study (Part II)

Previous epidemiological studies of miners occupationally exposed to radon have indicated a relationship between lung cancer risk and radon exposures. Radon exposures in Saskatchewan mines since 1975 are present but at much lower levels than in previous studies. The Saskatchewan Uranium Miner’s Cohort (SUMC) Study Group commissioned a study to assess the feasibility of an epidemiological study on these present day miners.

The statistical performance of two types of epidemiological studies, surveillance using a standardized mortality ratio (SMR) approach and a cohort approach using regression of lung cancer risk against radon exposure were considered. Exposure and demographic information from the National Dose Registry (NDR) were used to construct a representative cohort for the purpose of this feasibility study. This included development of a hypothetical cohort of individuals entering after 2000. The predicted number of lung cancers was calculated using baseline lung cancer rates, the age and sex distribution of the cohort, radon exposures and a relative excess relative risk model for the follow-up to 2010, 2020 and 2030. Simulation of the observed numbers of lung cancers, based on the predicted numbers, was used to assess the statistical performance of the two assessment models.

Occupational exposures were typically estimated to be small relative to lifetime residential exposures and also small relative to the likely variation in these residential exposures. Given the low occupational exposures and the risk model considered in this study, the modeled excess risk attributable to radon exposure was small compared to the expected, or baseline, lung cancer risk. With perfect information on exposures and correct baseline lung cancer rates, there is almost no chance that an epidemiological study would detect a statistically significant excess risk due to radon exposures. The selection of a baseline population for estimation of the expected lung cancers will be a challenge for the surveillance studies since small differences in smoking prevalence will likely overwhelm the potential risk attributable to radon exposures. Moreover, the occupational exposure to radon is thought to be small compared to the likely variation in exposure to residential radon. The results of a surveillance (SMR) study would be very sensitive to the choice of baseline lung cancer rates.

Operationally, a future study is feasible since data can be collected, methodology and measurement tools exist and there is experience in these types of analysis. From our point of view, future epidemiological study of these individuals will provide little scientific benefit to the understanding of the relationship between lung cancer risk and radon exposure. Likewise, a surveillance study (i.e., SMR approach) would have little statistical power to detect an excess risk due to radon exposure; however, it may be desirable to assess the potential for other diseases and to reassure miners and to demonstrate due diligence on the part of public health agencies and the companies.

Top of page

RSP-0179A – Technologies for mitigating tritium releases to the environment

Significant developments in tritium-capture technology have occurred over the past two decades. These developments are particularly well suited to the radio-luminescent industry. However, under the economic pressure of a limited market and increasingly more stringent safety and environmental regulations, the industry has been slow to incorporate those developments into their plants.

Tritium-capture technology has been developed for both air and inert gas streams. Tritium removal from air streams requires the oxidation of elemental tritium and the collection of the tritiated water on driers. Tritium recovery from inert streams can also use the oxidation. Alternately, metal getter technology can be used to detritiate inert gas streams by collecting elemental tritium directly without the oxidation step.

Direct elemental tritium recovery offers several advantages. Elemental tritium is significantly less hazardous to handle than tritium oxide. In some cases, the elemental tritium can be returned to the original storage device for reuse. On the other hand, oxidation technology tends to be more robust. However, unless special infrastructure is installed, oxidized tritium is a waste stream destined for land disposal. Each tritium-capture application should be carefully evaluated to assess the merits and drawbacks of oxidation against direct elemental recovery before selecting a capture technology.

Capture technology does not provide magical bolt-on solutions to process systems that suffer from chronic or repeated accidental releases. Rather this technology must be integrated with the operation of the process systems. Both the process systems and the capture technology need to be configured for reliable, fail-safe operation. For example, circulating closed-loop capture-subsystems are preferred to once-through subsystems. Process loop exhaust lines to the stack should be monitored for activity and interlocked with valves that automatically isolate the system from discharging tritium to the stack above predetermined limits.

Economics favours small-scale systems. Emissions should be treated as close to the source as possible and certainly before any significant dilution with air or inert gas occurs as the first step in reducing the size of the capture equipment. Distributed capture subsystems are preferred to one large clean-up system attached to the stack. The former costs significantly less. Additionally, releases in one process do not impact on the operation of other processes.

Top of page

RSP-0179B – Good work practice for effective tritium management

Tritium is the most benign of all radioactive elements. Its decay produces an electron with low energy that travels only 6 mm in air and can not penetrate the dead layer of skin. However, as an isotope of hydrogen, tritium easily exchanges sites with hydrogen in water and organic compounds. Attached to these chemical species, tritium can be easily incorporated in humans and in the environment. While a fatal exposure due to tritium uptake is highly unlikely, chronic exposure is likely unless precautions are taken to contain tritium.

With a modest investment in infrastructure, tritium can be contained and handled safely with negligible chronic and accidental release to the environment or exposure of staff.

This document develops an understanding of the fundamental nature of tritium and its interaction with materials. Factors that influence material compatibility with tritium include the stability of materials when subjected to low energy electrons and charged helium-3 particles resulting from tritium decay. Most metal containers prevent tritium from escaping unless they are heated above 150oC. Tritium readily permeates through elastomeric seals and vessels. These should be avoided in applications where airborne tritium or elemental tritium is present. Equations are provided to estimate tritium permeation through hot stainless steel surfaces.

Monitoring is a vital element of managing tritium exposures and releases. Real-time monitors provide immediate feedback to process loop operations and permit responses that limit both accidental and chronic exposures and releases to the environment. Passive monitoring is more sensitive than real-time monitoring. Its main function is to trend the contamination levels of a facility and the impact a facility has on its environment, and to demonstrate that emissions from a facility are within regulatory limits. Surface activity monitoring reduces the spread of contamination in a facility and exposure to employees. Smearing is a tool for estimating the radiological hazard of a surface. Surface activity monitors provide surface tritium inventory estimates for disposal of contaminated solid wastes.

Accidental and chronic emissions differ in magnitude of tritium release and in origin. Typically, tritium gas can escape through equipment or component failure or as a result of accidental evacuation of a process loop. Secondary containment with attendant and simple recovery systems is very effective in mitigating the escape of tritium. These systems should circulate the gas through the recovery system. Evacuation lines can be fitted with automatic isolation valves that are coupled to in line process monitor to prevent the accidental evacuation of tritium. Chronic releases can be reduced through the use of simple, once-through tritium collectors. The cost of installing tritium recovery equipment must be balanced against magnitude and the risk posed by the release.

Facilities should be sectored into three zones. Zone 1 is intended for inactive work. Zone 2 is a transition between Zones 1 and 3 where some contamination may occur but no open sources are present. All active work is executed in Zone 3. Air flows should enter the facility through Zone 1 and leave via Zone 3.

Top of page

RSP-0180 – The toxicity investigation of uranium to aquatic organisms

The acute and chronic toxicity of uranium to six freshwater species in waters of low alkalinity (~5 mg/L as CaCO3) and various harnesses (5 to 240 mg/L as CaCO3) were investigated in this study.

There were no apparent effects of hardness on the toxicity of uranium to fathead minnows (Pimephales promelas) in 7-d survival and growth tests. Tests that were conducted at mean measured water hardnesses of 23, 72, 131 and 244 mg/L as CaCO3 yielded 7-d LC50 values (in mg U/L with 95 percent confidence intervals; based on geomean measured uranium concentrations) of 1.6 (1.5 – 1.8), 2.1 (2.0 – 2.1), 2.0 (2.0 – 2.1) and 1.5 (1.3 – 1.7) respectively. There was no significant effect on growth at the concentrations tested.

Rainbow trout fry (Oncorhynchus mykiss) were not sensitive to uranium toxicity in 96-h static tests, and no effect of hardness was observed. Tests that were conducted with water hardnesses of 20, 68, 126, and 243 mg/L as CaCO3 yielded similar 96-h LC50 estimates of 4.2 (2.6 - 6.7), 3.9 (2.4 - 6.3), 4.0 (2.5 - 6.3) and 3.8 (2.4 – 5.9) mg U/L, respectively.

Two embryo/alevin (EA) tests were conducted with rainbow trout embryos at two mean measured water hardnesses: 6 and 61 mg/L as CaCO3. The exposure period was 31 and 30 days, respectively, from day of fertilisation. Toxicity was significantly higher in the softer water, with an EC50 estimate of 0.46 (0.40 – 0.51) mg U/L. The EC50 estimate in the harder water was 0.64 (0.63 – 0.65) mg U/L.

Three-brood survival and reproduction tests with Ceriodaphnia dubia were conducted at four mean measured water hardnesses: 5, 17, 124 and 252 mg/L as CaCO3. In all tests, there was 100 percent mortality in the highest test concentration, 0.40 – 0.46 mg U/L, within 24–96 h of test initiation, with complete to partial mortality generally occurring in the next highest concentration, 0.16 - 0.18 mg U/L, by test termination. The LC50 estimates were not different among the tests, and were 0.16 (0.12 – 0.17), 0.14 (0.12 – 0.18), 0.10 (0.014 – 0.20) and 0.11 (0.068 – 0.21) mg U/L in tests with water hardnesses of 5, 17, 124 and 252 mg/L as CaCO3, respectively. Reproductive toxicity appeared to be lower at the mean measured water hardness of 17 mg/L as

CaCO3, compared to the other water hardnesses. The IC50 estimates for reproduction were 0.092 (0.079 – 0.10), 0.11 (0.11 – 0.11), 0.053 (0.036 – 0.098), and 0.095 (0.062 – 0.11), respectively.

Survival was a more sensitive endpoint than growth for Hyalella azteca in 14-d survival and growth water-only tests. The effect of hardness was apparent as toxicity was reduced with increasing hardness. The 14-d LC50 estimates were 0.017 (0.0088 – 0.038), 0.14 (0.12 – 0.16), 0.20 (0.17 – 0.24) and 0.34 (0.19 – 1.8) mg U/L, in mean measured water hardnesses of 17, 61, 123 and 238 mg/L as CaCO3, respectively.

The toxicity of uranium to Selenastrum capricornutum was investigated using the 72-h growth inhibition test with the microplate technique. Tests were conducted at five nominal water hardnesses of 5, 15, 64, 122 and 228 mg/L as CaCO3. The IC50 estimates were 0.16 (0.11 – 0.19), 0.17 (0.078 – 0.20), 0.10 (0.079 – 0.11), 0.20 (0.11 – 0.30), and 0.20 (0.16 - 0.21) mg U/L, respectively.

Lemna minor was not sensitive to uranium toxicity in the 7-d growth inhibition test, although a slight effect of hardness was observed. Tests conducted with a mean measured water hardness of 35 mg/L as CaCO3 produced IC50 estimates (frond number) of 7.4 (6.4 – 9.2) mg U/L, whereas tests at 137 mg/L as CaCO3 produced IC50 estimates (frond number) of 16.4 (14.8 – 18.2) mg U/L (based on time-weighted mean uranium concentrations). Frond number yielded lower estimates than dry weight. The IC50 estimates (dry weight) were 13.1 (9.1 – 15.6) and 35.5 (5.6 – 53.2) mg U/L for the mean measured water hardnesses of 35 and 137 mg/L as CaCO3, respectively.

In general, H. azteca, C. dubia, and S. capricornutum were the most sensitive species to uranium toxicity with toxicity estimates that ranged between 0.017 and 0.34 mg U/L. Rainbow trout embryo/alevins were the next most sensitive, with estimates between 0.46 and 0.64 mg U/L. The 96-h lethality of uranium to fathead minnows was almost half that to rainbow trout fry (1.8 – 2.1 mg U/L compared to 3.8 – 4.2 mg U/L). L. minor was the least sensitive species tested, with toxicity estimates starting at 7.4 mg U/L. A decrease in toxicity with an increase in water hardness was only observed in the H. azteca, L. minor and rainbow trout embryo/alevin tests. This trend of increasing toxicity with increasing hardness was most pronounced in the Hyalella tests where the 14-d LC50 estimate increased from 0.017 mg U/L in soft water to 0.35 mg U/L in hard water, a decrease in toxicity by a factor of 20.

Top of page

RSP-0181 – Survey of CANDU fuel bundle experiments under high temperature conditions

Severe degradation of CANDU fuel cooling has been postulated for a number of accident scenarios within the design basis. The scenarios include the early stages of a large loss of coolant accident (LOCA), some single channel events such as breaks in inlet feeders and certain loss of flow transients. As a result, the fuel temperatures may exceed the threshold for onset of thermally induced deformations and a host of other related high temperature phenomena. This may contribute to challenge the ability of the fuel bundle to maintain coolable geometric configurations, which do not threaten the integrity of the whole channel. The severe mismatch between energy deposition in the fuel and removal of heat from its surfaces, exasperated by a positive void reactivity induced power pulse early in the large LOCA transients, results in a combination of interrelated high temperature phenomena (e.g., fuel sheath strain, non-uniform fuel dimensional changes, element sag, end plate deformation, zircaloy oxidation, melt relocation) and can significantly alter the bundle and pressure tube geometries. While taking extreme measures to minimize the likelihood of such accidents, licensees of CANDU power reactors must convincingly demonstrate that any credible challenges to fuel and channel integrity are handicapped by both the fuel design as well as the engineered safety systems. Licensees must also demonstrate that the fuel can maintain a coolable geometry during the heatup transient and following reflood. Severe fuel bundle heatup and deformation has also been reported in licensing submissions for LOCA scenarios involving an additional failure to inject emergency core cooling in a timely manner.

For large LOCAs, an early pressure tube rupture caused by contact with an overheating and deformed bundle is an undesirable outcome that must be precluded by integrated and unambiguous analyses. These analyses must use appropriately validated computer codes exercised over the whole range of anticipated accident conditions, representing to the fullest extent possible the most relevant underlying phenomenology. For accidents that combine a LOCA with a loss of emergency core coolant (LOECC) injection, high bundle temperatures and deformations are inevitable and uncertainties related to hydrogen and fission product source terms can be minimized with computer codes that recognize and model the various thermal, chemical, metallurgical and mechanical changes which define the state of the deformed bundle and its ability to transfer heat to the available heat sinks and contain the fission products. Validated high temperature fuel bundle behaviour computer codes are also required for seemingly benign slow transients such as degradation of single channel flows, loss of a pump, etc., which may cause selective fuel dryout at a few locations within the fuel string.

This report examines the experimental database available to provide insights into phenomenology of high temperature behaviour of full size fuel bundles and to help validate available computer codes used in licensing submissions and in other supporting analyses. It focused on experiments that establish the thermal threshold for fuel element deformations and on published high temperature experiments with both authentic and simulated fuel bundles. It reviewed fuel bundle behaviour related computer codes used in licensing submissions over the last three decades. The review was not expected to be a repository of all fuel bundle related experiments or of computer codes that may somehow be still under development, and did not attempt to do so. It did result, however, in the replication of an exhaustive list of related literature that can be consulted to further investigate the issues and to confirm the conclusions arrived at in this study.

This review concludes that significant and serious deficiencies exist in the range of fuel bundle geometries and phenomena that the high temperature fuel bundle computer codes are able to model, and in the experiments undertaken to validate them. It recognizes the limited diversity, limited abilities and limited applicability of the currently available simulation methods and recommends a modelling approach that allows an integrated representation of fuel channel behaviour with due consideration of all important phenomenological feedbacks and an ability to model a wider range of fuel behaviour so that the uncertainties coming sometimes as a surprise to the industry, such as an increased initial power pulse for large LOCAs do not handicap the safety analysts in their ability to analyze fuel transients without undue conservatism and without having to wait for years for new methodology.

While advances have been made in thermal modelling of under formed bundles within integrated thermal-hydraulic codes such as CATHENA and TUF, there have been no new computer codes suitable for modelling the complex phenomenology of fuel bundle heatup and deformations developed for over a decade, except for some feeble attempts in modelling full fuel bundles by recursively using existing computer codes (e.g., ELOCA) repackaged under new names (e.g., FACTAR) and dissociated from core thermal-hydraulics. It is difficult to tell what is driving the lull in the research and development in this field, but the results are obvious. Even simple attempts to combine thermo-mechanical behaviour modelling capabilities of ELOCA with fission product release behaviour of SOURCE have slowed down. Experimental programs in the field have been essentially terminated in spite of an increased awareness of code uncertainties in prediction of degraded bundle behaviour.

This report proposes an architectural outline of a new computer code to predict high temperature behaviour of a fuel channel. The difference between this approach and the existing models is that it would recognize and dynamically utilize the feedback of behaviour in other bundles, other channels and heat transport system response, as well as model many additional phenomena. Such an approach has been utilized in an existing multi-channel, severe core damage code MAAP-CANDU for which we created the CANDU core component models back in the days of 80486 computers. Given the exponential growth in computational power and advent of new numerical techniques, it is entirely possible to develop such a code within one year.

The overall conclusion is that the industry must apply resources, commensurate with the uncertainties and risks posed by accidents that may lead to fuel bundle deformations, towards timely development of validated computer codes that better represent anticipated behaviour of actual fuel bundles under highly degraded cooling conditions resulting in high temperatures. The ultimate benefit may be to the utilities required by regulators now to derate and operate with what may actually turn out to be undue conservatism because tools do not exist to reduce uncertainties. In addition, an increased understanding of the high temperature fuel behaviour, inevitable in a number of postulated accident scenarios within the design basis, may lead to design and operational changes that better serve public safety. Given that a number of fundamental safety issues have remained open over the last decade, the regulators need to insist on more stringent time lines on development, delivery and use of properly validated state-of-the-art accident analysis computer codes.

Top of page

RSP-0182A – Designing and implementing a radiobioassay program

This report updates the information given in a previous report by the Federal-Provincial Working Group on Bioassay and in Vivo Monitoring Criteria, entitled "Bioassay Guideline 1 – General Guidelines for Bioassay Programs" (Health 1980). It presents the basis for designing and implementing a radiobioassay program as part of the internal dosimetry component of an occupational radiation protection program. It answers three questions:

a) Is bioassay necessary? b) How and how often should it be done? c) What do the results mean?

Under certain employment conditions, individuals may be exposed to radionuclides in a variety of chemical forms that could be inhaled, ingested or absorbed through intact skin or open wounds. The purpose of a radiobioassay program is to monitor the intake of radioactive materials by workers. Monitoring results are used to estimate equivalent or effective dose to the affected individuals. Thus a radiobioassay program also provides assurance that the internal radiation control program in the workplace is functioning.

The report was prepared by the Working Group on Internal Dosimetry of the CNSC. The working group is comprised of experts in the field of internal dosimetry from industry, health sciences and government.

Top of page

RSP-0182B – Radiobioassay and dose assessment for intakes of tritium

The purpose of this report is to update previous information present in documents such as the Report of the Working Group on Bioassay and in vivo Monitoring Criteria (Health and Welfare Canada, Environmental Health Directorate, 82-EHD-87). It incorporates the latest information on radiobioassay for tritium from national and international sources.

This report offers a comprehensive treatment of all aspects of radiobioassay for tritium. It presents a method for selecting workers who should participate in a tritium radiobioassay program based on the quantities handled. The important forms of tritium exposures are identified and monitoring frequencies and methods are recommended. For the major radiobioassay methods, i.e., urinalysis, the calibration and quality aspects of the liquid scintillation analytical system are addressed.

An important component of this report is the consideration of metabolic models for various forms of tritium and their use in assessing dose resulting from tritium exposures. The interpretation of tritium monitoring results and dose assessment is illustrated by worked examples for chronic and acute exposures. The accuracy in dose assessment that can be achieved for a 14 day monitoring period for tritium oxide is also discussed.

Top of page

RSP-0183 – Background radionuclide concentrations in major environmental compartments of natural terrestrial ecosystems

The Nuclear Safety and Control Act mandates the CNSC to protect the environment from radioactive materials. The CNSC must ensure licensed facilities and their operations do not cause unreasonable risks to the environment. The assessment of potential effects of radionuclides released to the environment still requires more reliable background data on radionuclide concentrations in natural terrestrial ecosystems. Data are also lacking on the assumptions of secular equilibrium of the radionuclides in the thorium (Th) and uranium (U) decay chains. Further, few data collected to establish background concentrations in terrestrial media (soil, plants and biota) have been collected in an integrated manner so that radionuclide transfer between various compartments of the terrestrial ecosystem can be assessed for a given site. The purpose of this project was:

  • to provide an integrated sampling methodology for future use in collecting terrestrial background data to fill knowledge gaps for ecological assessments
  • to carry out integrated sampling in key field locations to test the sampling methodology and to begin to provide current background data for U and Th series radionuclides (and selected metals), which will implicitly include information on the transfers of these nuclides among terrestrial biosphere compartments

The collection of environment samples appears simple. The collection of meaningful environmental samples that will provide data to answer specific questions is very complex. Subtle nuances in the statement of purpose profoundly affect the methodological details.

The present project developed a sampling methodology to ensure the project met its objectives. A data quality objectives (DQO) approach was followed. Analytical cost and representation of geographic domains were found to be the predominant considerations, and the resulting methodology was a stratified sampling scheme that maximized cost/benefit within the expectations specified for this project. In addition, the next level of methodology developed was the field instruction, which included criteria to: define the size and inclusiveness of a site, select the species to sample and determine timing of sampling, and specifications for: the handling of samples, ancillary information, cleanliness, archiving and replication. It is anticipated that with this document, a skilled environmental technologist (or equivalent) would be able to collect samples in a manner that is consistent with the overall present and future program objectives and methods.

Samples were collected at seven sites incorporating several major physiographic regions and ecozones, listed below approximately east to west:

  • Avon Valley in Hants County of Nova Scotia, an area with known elevated U concentrations (so that secular equilibrium could be investigated) in the Appalachian physiographic region, Atlantic maritime ecozone, with podzol soils
  • Stormont County of Ontario, in the St. Lawrence lowlands physiographic region, mixed wood plains ecozone, with grey-brown luvisol and related gleysol phase soils.
  • Leeds County of Ontario, in the St. Lawrence lowlands physiographic region, mixed wood plains ecozone, with calcareous brunisol soils
  • Bancroft area of Ontario, in a region with elevated U concentrations and where there may be future licensed decommissioned mine sites, in the Canadian shield physiographic region, boreal shield ecozone, with Podzol soils
  • Black Lake in Manitoba, in a region with elevated U concentrations, in the Canadian shield physiographic region, boreal shield ecozone, with organic soils
  • Pinawa Channel in Manitoba, in the Canadian shield physiographic region, boreal shield ecozone, with heavy clay grey luvisol soils
  • Milner Ridge in Manitoba, in the Canadian shield physiographic region, boreal shield ecozone, with acidic Brunisol (some marginal Podzol) soils

The sites were carefully characterized and documentation provided on geology, soils and vegetation. From these sites, 162 plant samples representing 38 species were collected. They were chosen to represent the dominant species present, and to include species sampled at other sites (for inter comparison), species that have fruit or are otherwise edible, and species from a range of plant types including fungi. Plant species, using common names, include a considerable breadth of species: apple, balsam fir, blueberry, buckthorn, cedar, chokecherry, clover, cranberry, eastern hemlock, fern, grape, grass, hawthorn, hazel, juniper, Labrador tea, lichen, locust, maple, milkweed, moss, mushroom, oak, pine, prickly ash, raspberry, rose, Saskatoon, sarsaparilla1, snake grass, spruce, strawberry, sumac, vetch, willow and wintergreen. Eighteen of these species have been sampled on more than one site within the sampled ecozones to facilitate quantifying spatial variability. All soil and plant samples were analysed for one of three suites of radionuclides and for a multi element analysis by inductively couple plasma mass spectroscopy (ICP MS). Some samples were divided into tissues (e.g., leaf versus twig) and the tissues analyzed separately by ICP MS.

A literature review of Canadian terrestrial biosphere background concentrations, with emphasis on natural (primordial) radionuclides in the U and Th decay series in natural settings was also completed. Of particular importance is information on the occurrence or degree of secular equilibrium; in particular, the isotopic activity ratios of 210Po/210Pb and 210Pb/226Ra. The mean and range of background soil activities of natU, 226Ra, 210Po and 210Pb in all regions are similar, suggesting that they are near isotopic equilibrium and that there has been long term buildup in the soil. However, the isotopic ratios of 210Po/210Pb and 210Pb/226Ra for all vegetation over all regions are highly variable, with certain trends. The data generally indicate that 210Po/210Pb ratios are typically less than 1.0, indicating that there has been not been sufficient time for ingrowth to reach secular equilibrium. Coniferous trees generally have 210Po/210Pb ratios closer to 1.0, indicating that retention on needles and bark contributes to a longer time for ingrowth than on annuals such as grasses and forbes and deciduous shrubs. Lichens also have 210Po/210Pb ratios closer to 1.0 as dust on lichen can have a ~50-year retention half-life so equilibrium is more likely. Wetland species generally have much lower 210Po/210Pb ratios, likely due to the decreased influence of atmospheric deposition from radon (222Rn) emanating from soils.

While 210Pb and 210Po were in near secular equilibrium in plant tissues, they were not in equilibrium with the parent 226Ra. The 210Pb/226Ra ratios deviate positively from equilibrium, suggesting atmospheric deposition of 210Pb from radon decay. Annual plants appear to express a lower 210Pb/226Ra ratio while perennials express higher ratios, also confirming the suspected important pathway for 210Pb. The high variability of this ratio in vegetation, ranging from 0.28 to 391, with the highest values generally attributed to lichens and coniferous trees, suggest that this parameter may not be particularly useful in categorizing class or species behaviour. The activities of natU, 226Ra, 210Po and 210Pb were summarized for all classes and species of vegetation and are generally much lower than soil activities.

The results from the samples collected in this study demonstrate a number of expected trends, and allow for some additional interpretation. The key points are listed here as bullets:

  • Detection limits for the natural radionuclides and 137Cs, although the best available, were not sufficiently sensitive to detect these radionuclides in all samples. This is despite collection of very large plant samples, which allowed ashing as a pre concentration step in the analysis.
  • Duplicate analyses by the radiochemical methods were satisfactory, but the agreement between duplicates was probably diminished by the fact that all values where near their detection limits (where by definition errors are 50 percent). Duplicate analyses by the ICP MS methods were very good.
  • The seven sites sampled had a very broad range in the key soil properties. However, the radionuclide concentrations were not very different among the sites or among the soils within a site.
  • Secular equilibrium was found in soil between 210Po and 210Pb, and between 228Th and 232Th. However, activity concentrations of 210Pb were about 2.5 fold above that of the parent 226Ra, and were perhaps more elevated in the topsoil that the subsoil. The excess 210Pb was attributed to the deposition of 210Pb onto the topsoil from the decay of volatile radon.
  • Secular equilibrium was not found in plants, the 210Po/210Pb activity ratio had a median of 0.6, 210Pb/226Ra had a median of 14 and 228Th/232Th had a median of about 4. For the 210Po/210Pb activity ratio, this was attributed to insufficient time for ingrowth, for the 210Pb/226Ra ratio this was attributed to deposition onto plants of 210Pb from the atmosphere and for the 228Th/232Th ratio this was tentatively attributed to an effect of alpha recoil in enhancing the bioavailability of 228Th.
  • Fruit, cones and pods had lower concentrations of elements than did the corresponding leaf and twig samples. Bole wood from trees was especially low. Lichens and moss had high concentrations, reflecting that they retain dust very effectively.
  • Plant/soil concentration ratios (CR) varied significantly among sites and among plant types.
  • There were distinct differences in CR depending on whether the data were from radiochemical or ICP MS methods. The differences were perhaps explainable, related to the role of atmospheric radon in delivering 210Pb to plants and the possible role of alpha recoil making 228Th more bioavailable than 232Th.
Table: The decisions defined by the DQO process, and the abbreviated answers are:
Are concentrations in one compartment positively correlated to concentrations in the corresponding substrate media? With respect to plants and soils, the concentrations are positively correlated but the relationship is not linear.
Do the concentrations, or the transfers among compartments, of the nuclides vary among sites; and is this variation related to soil properties? Concentrations and concentration ratios vary among sites, and this variation is related to soil properties. Specifically, acidic sandy soils tend to have enhanced transfer from soil to plants.
Are the activity ratios among radionuclides in a decay chain relatively constant within or among sites? The 210Po/210Pb activity ratio is fairly constant among sites, but the other ratios examined were very variable.

Overall, the study has established a sampling protocol, has reviewed previous data and has provided new data that contribute to the characterization of background concentrations of primordial radionuclides in natural settings.

1 Also called Sarsparilla or Sasparilla. The spelling "Sarsaparilla" was taken from a Manitoba Agriculture website, (province of specimen’s origin).

Top of page

RSP-0184A – Licensing guide: Design

This project was established to develop licensing basis documentation for the ACR 700 reactor that will be used by CNSC staff as a guide for the assessment of licensability of the ACR design. The project has three deliverable documents: Report on Early Identification of Issues, Licensing Basis Review Guide and ACR Licensing Guide: Design. The approach taken to develop these documents will be top-down, systematic and comprehensive. Current regulatory requirements and industry standards and practices for the licensing of a CANDU reactor have been examined and the suitability for application to the ACR assessed. Where necessary, changes are proposed and/or new requirements recommended. The IAEA Safety Standards Series Document NS-R-1 entitled Safety of Nuclear Power Plants; Design is used as the template for the Basis for the Licensing Guide: Design.

The report proposes modifications that will make the overall licensing process more risk informed than the current deterministic based approach. It requires a combination of deterministic analysis and probabilistic safety assessments. The report recommends the adoption of quantitative safety goals, and a new event classification scheme for analysis of accidents is introduced. Recommendations are also made to change several of the current rules for the design of systems in the areas of reliability, shutdown requirements, trip requirements, sharing of instrumentation and equipment between process and safety systems, safety classification, containment leakage requirements and the introduction of operating limits and conditions. These modifications will bring the Canadian licensing process more into line with accepted international practice; at the same time ensuring plants built to these requirements will provide a high level of safety.

As far as is practicable this document proposes requirements that can be applied to both future CANDU and future non-CANDU reactors.

Top of page

RSP-0184C – Review of licensing approach for advanced CANDU reactor

Atomic Energy of Canada Ltd. (AECL) is developing a new Candu reactor design called the Advanced Candu Reactor (ACR). The CNSC is undertaking a pre-licensing review of the reactor design. The objective of the review is to issue a formal statement as to whether the design is licensable in Canada or under what conditions it would be licensable.

An important component of the pre-licensing review is the development of a Licensing Basis Guide, whose purpose is to identify all the licensing requirements and acceptance criteria that would apply to the ACR. These will be based on Canadian and international regulatory standards and guides, and on documented regulatory and industry practice and experience.

The bulk of the licensing basis development work is being performed by a team of consultants led by R.A. Brown and Associates Ltd. (RABA) under CNSC contract. The recommendations of the RABA consultants are expected to form the basis of a Licensing Basis Guide to be issued by CNSC staff.

The present document is the product of a small separate project commissioned by CNSC staff, to assist CNSC staff in comparing the safety philosophies underlying the current regulatory framework for power reactors in Canada and the new proposals being developed by the RABA consultants.

In this project, design documentation submitted by AECL to the CNSC (references 1,2,3) has been reviewed as appropriate, as well as documents used by AECL at familiarization sessions with CNSC staff. Reports prepared by the RABA consultants (references 4,5,6,7,8,9,10) were also reviewed in accordance with the terms of the contract. While references 11, 12, 13, and 14 were not formally part of the contract, they were reviewed because of their relevance or because they were updated versions of the earlier reports.

It is not the purpose of this review to undertake a detailed review of either the ACR design or the RABA proposals. Nor is it within the scope of this project to make alternative proposals with respect to the licensing requirements for nuclear power plants in Canada. The objective is to address key differences between the current Canadian regulatory regime and the approach proposed by RABA, and to comment, on the basis of the author’s experience in nuclear reactor regulation, on the safety implications of the proposed changes.

Top of page

RSP-0185 – Radiobioassay and dose assessment of mixed fission and activation products

This report describes the components of an effective bioassay monitoring program for intakes of fission and activation products. It also provides examples on how such a program may be implemented, addressing measurement methods and dose estimation. The best available metabolic and dosimetric models currently available have been used in the preparation of this report.

Addressing each radionuclide in every intake scenario in a rigorous manner would result in a large and cumbersome document that would have little value to a general audience. The subject is therefore addressed using as examples 60Co, 147Pm, 239Pu, and 90Sr, which are dealt with in detail. Applying the techniques presented in the worked examples involving these radionuclides and referring to Table 2a and Table 2b should enable the reader to derive the values of interest for other radionuclides.

Work on this report was begun by the now defunct Federal Provincial Working Group on Bioassay and In Vivo Monitoring Criteria. The final document was drafted by the CNSC Working Group on Internal Dosimetry and revised by J. R. Johnson under a contract to IDIAS, Inc.

Top of page

RSP-0186 – Safety analysis review guide for non-power reactors

This is a Safety Analysis Review Guide for Non-Power Reactors. It provides practical guidance and instruction in preparing the Safety Analysis Report (SAR) for Non-Power Reactors in Canada. It describes the necessary content of the SAR in order to meet the IAEA Safety Requirements set out in IAEA-DS-272. It also provides a selection of worked examples in various degrees of rigor to illustrate how the analysis methods could be commensurate with a graded approach to meeting the requirements in IAEA-DS-272. The graded approach is a risk informed approach to ensuring that the design and operation of a non-power reactor has adequate defence in depth to meet fundamental safety criteria, where the disposition of postulated events is in a manner commensurate with the risk of the facility. It is used to help determine the requirements for the SAR.

Top of page

RSP-0187 – Creation of combined dose records for the Cameco study cohort

A Joint Federal/Provincial Panel on Uranium Mining Developments in Northern Saskatchewan, appointed in 1991, issued a report in October 1993. This report recommended the undertaking of an epidemiologic study of all Saskatchewan uranium miners, past, present and future. As a result, the CNSC has taken the lead in a study, with financial support from both the Saskatchewan Government and industry. This study includes updated mortality and exposure estimate data for former miners at Beaverlodge, Port Radium and workers at Port Hope.

The purpose of this work is to produce a comprehensive file of dose records for the Cameco cohort by combining early dose records from Cameco with those of the National Dose Registry (NDR). The use of the NDR’s records add non-Eldorado dose records that were not available to Cameco, thus providing more complete dose histories. Health Canada’s National Dose Registry (NDR) agreed to provide data for the years 1980 to 2000 for the study in return for the early Cameco (Eldorado) dose records that are not currently in the NDR.

The resulting file will be used in an epidemiological study of Cameco workers and for updating the missing early dose records in Health Canada’s NDR.

The work was broken down into three major Parts:

PART 1: The "Non-Eldorado Dose" Linkage

The linkage of a file of workers with some early, non-Eldorado doses from Cameco to the NDR identified the cases not in the NDR. These workers’ doses were then calculated by Mr. S. Frost at Cameco.

PART 2: The linkage of the complete CAMECO nominal roll to the NDR

This linkage was used to establish the cross referencing between records in the nominal roll and the NDR for use in combining and comparing the NDR and Cameco dose records.

PART 3: The production of a comprehensive dose file for the CAMECO cohort

The linking of the cohort dose records from Cameco to those in the NDR to produce a dose file for the principal investigator, Dr. G. Howe to use in the epidemiological study of uranium miners. A secondary file was also produced to include the missing early Eldorado dose records in Health Canada’s NDR.

Top of page

RSP-0188 – Summary report of the Eldorado nuclear cohort study

This report describes the record linkage and basic data quality work carried out at Statistics Canada for an epidemiological mortality and cancer study of 19,855 Cameco/Eldorado Nuclear employees. The overall study is an update of an earlier retrospective (1950-1980) mortality study of Eldorado Nuclear workers employed at the Port Radium, Beaverlodge and Port Hope locations. This is the first time that an "alive" follow-up was completed.

The following record linkages have been carried out: (1) an internal linkage of the cohort to investigate and remove duplicate records for the same employee; (2) an "alive" follow-up to determine the vital status of individuals in the cohort; (3) a linkage to the Canadian Mortality Data Base (CMDB) for event years 1950 to 2000 inclusive; and (4) a linkage to the Canadian Cancer Data Base (CCDB) for event years 1969 to 2000 inclusive.

The first step in the process flow was to run a series of programs against the cohort file to identify data gaps and anomalies, and to standardize name and date variables. Then, an internal linkage was carried out on the cohort file to identify duplicate records. After the manual resolution of the internal linkage was completed by members of the study team, an alive follow-up was carried out to determine the vital status of the individuals at the end of the follow-up period. This activity also served as an aid in the identification and confirmation of deaths and as an effective tool for evaluating the mortality linkage. This evaluation was possible for the 60 percent of individuals who had a social insurance number (SIN) either supplied or found through the use of a composite key generated as part of the work process. On the original cohort file, 43 percent of the records had a valid SIN supplied, while the exact key match generated a SIN for another 17 percent of individuals. Records which matched on social insurance number were then confirmed using other identifying information, such as names and birth dates.

Following the alive follow-up, a mortality linkage was carried out using the CMDB for the years 1950 to 2000 inclusive. As is the convention in probabilistic linkage, it was necessary to define the item(s) used as a pocket to reduce the number of records compared and pairs created for the linkage. Only records from each file containing the same pocket value were cross-matched against each other.

The items compared on the cohort and mortality file to determine if they were the correct individual included the following: surname, given names (first and second), sex code, birth date, birth place, death date (if available), geography (including postal code and province of residence), parent’s initials and place of birth for parents. These items were compared, when available, on the two files. The date that the person was last known alive was also taken into account, unless a death was recorded, in which case the date of death was compared.

Of the 19,855 individuals in the cohort, potential linkages to the 1950–2000 CMDB were reviewed. During the manual resolution phase, physical copies of death registrations were consulted when the linkages were uncertain. After manual review, 5,928 death links (5,729 males and 199 females) were accepted. An additional 46 links (all males) were found through manual searches. There were 32 deaths identified in the alive follow-up that could not be found through the mortality linkage. These could potentially be individuals who died outside of Canada.

From the alive follow-up and mortality linkage work, 8,218 individuals (41.4 percent) were determined to be alive at the end of the follow-up period, 5,974 (30.1 percent) were confirmed dead, and 5,631 (28.5 percent) were lost to follow-up prior to the end of 2000.

Finally, a cancer linkage was carried out using the CCDB for the years 1969 to 2000. The pockets and rules used for this linkage were similar to those used for the mortality linkage. After manually resolving the potential links, there remained 2,469 accepted links to the CCDB.

Top of page

RSP-0189 – Review of ACR-LBD-001, licensing basis document for new nuclear power plants in Canada

The purpose of this project was to assist the staff of the CNSC in its development of a new licensing basis document for nuclear power plants in Canada by commenting on the document itself and the input used for its generation.

In this report, the contractor’s comments on the LBD were to be made in the light of: Canada’s current safety approach for CANDU type power plants; the CNSC’s current approach to licensing nuclear power plants; recommendations and rationale made by R. A. Brown and Associates (RSP-0184A and RSP-0184B); and, the International Atomic Energy Agency’s Safety Standard NS-R-1, Safety of Nuclear Power Plants: Design. In particular, the contractor was to comment on whether the new licensing guide: is a reasonable evolution of the approach outlined in AECB 1059, Reactor Licensing and Safety Requirements, and subsequent elaborations; whether proposed changes in the approach are a reasonable departure from the current approach to nuclear power plant safety and licensing; and whether, on balance, the new approach represents an advance in the current approach to nuclear power plant safety and licensing.

The opinions expressed in this report are those of the contractor’s based on his experience. Any decisions or actions characterized in my report as questionable or as mistakes, particularly in the historical review, reflect his opinion of those decisions or actions. The contractor has tried to be objective but even a sense of objectivity is relative. He has tried to avoid understatement, for fear that the message might be lost, and overstatement, for fear that the message might be dismissed.

Top of page

RSP-0190 – Independent review of current status of leak-before-break assessment at Darlington NGS

Clause 25 of the construction licensing requires that the primary heat transport system shall be restrained to the extent necessary to ensure that failure could not cause consequential damage that would invalidate the analyses in the safety report." Instead of providing massive pipe-whip restraints to do this, the licensee proposed a new approach by demonstrating that even if a through-wall crack were to develop, they would detect it well before a catastrophic rupture of the piping could occur. This approach, based on fracture-mechanics, is referred to as leak-before-break, LBB.

References [1] and [2] contain the licensees arguments based on LBB, that the DNGS large-diameter PHT pipes meet the licensing requirements for unrestrained piping. OPG wish to close out the outstanding regulatory commitments on clause 25.This report reviews OPG’s submissions in references [1] and [2] and recommends that they meet the intent of clause 25 and that the remaining action items can be closed.

OPG’s reliance on leak-before-break (LBB) for Darlington was the first and only time that a licensee has used the technology in support of application for a license. CNSC staff has expressed its views on LBB in memos, letters, and license conditions. This report recommends that the CNSC needs documents that are more formal.

In the licensees submission there is no provision of ongoing evidence from operating experience that the probability of catastrophic rupture remains low.

Top of page

RSP-0191 – Noble gases in Canadian Shield groundwaters at Yellowknife, N.W.T.

The CNSC maintains an interest in the composition and origin of Canadian shield groundwaters found at depths up to and exceeding 1000 m, and their relevance to the safe disposal of nuclear waste in such an environment. One geochemical tool of interest is the concentration of noble gases in deep Shield brines, which can be of atmospheric and geogenic origin. Their concentrations provide information on the recharge history and age of these fluids. This study of noble gas concentrations and ratios was carried out by the University of Ottawa, funded by a research contract with CNSC (contract no. R177.2). The research program developed from a Phase I program (Clark et al., 2002), which involved the design and building of a noble gas extraction line with quadrupole mass spectrometer analyser. This Phase II study involved improvements to the noble gas measurement methodology and application to a final round of samples collected at the now closed Con Mine, Yellowknife.

Top of page

RSP-0192 – Guidelines for safety assessment of application using best-estimate and uncertainty analysis methods for CANDU nuclear power plants (revision 1)

Guidance is provided for the review and licensing of best-estimate safety analysis methods and methodologies which are used to calculate CANDU nuclear power plant response for postulated design basis accidents. The primary focus of the review of best-estimate methods starts with the development of a phenomena identification and ranking table of PIRT. The PIRT identifies the key phenomena which are most important for modeling a postulated design basis accident. The PIRT also indicates the phenomena that the computer code must model to accurately perform the plant analysis with confidence. Guidance is provided on the development of the PIRT and how this aids in the development of the computer code validation matrix for both separate effects tests as well as integral effects tests.

The computer code validation effort is a key portion of the best estimate methodology since the development of the computer code uncertainty and the associated ranging of these uncertainties are a result of the validation effort. The validation effort also identifies possible computer code bias which should be eliminated or minimized as much as possible. Guidance is provided on how the computer code validation should be performed.

The plant state must also be specified and the associated uncertainties with the plant state must also be identified. Examples of plant states for a PWR analysis is given with guidance on how sensitivity calculations can be performed to rank the plant initial condition and accident boundary condition uncertainties.

The different methods of uncertainty combination are discussed including the response surface approach methods that have been used as well as the ordered statistics approach developed using the Wilks’ methodology.

Lastly there is discussion and guidance on how to maintain compliance for different operating plant conditions using a best estimate methodology for normal and off-normal conditions.

Reponses to Canadian industry comments on the report are included in an appendix as well as the questions asked for Westinghouse and Framatome-ANP for their best-estimate computer codes and methodology.

Top of page

RSP-0193 – Radon progeny exposure and lung cancer: A mortality study of Newfoundland fluorspar miners 1950–2001

This study is an updated analysis of the mortality experience of a cohort of Newfoundland fluorspar miners exposed to radon progeny. The mortality experience of the cohort between 1950 and 2001 was determined by linking occupational records to the Canadian Vital Statistics database. Our findings are based on a total of 328 miners who worked exclusively on the surface, and 1,742 individuals that were exposed to radon from working underground. All miners were male. Annual exposure to radon, in working level months (WLMs), was estimated for each cohort member. In 1978 mining operations ceased, and no occupational exposures were assigned to the cohort after this time.

When compared to Newfoundland males, an increased number of deaths were observed among the fluorspar miners for lung cancer, silicosis, and accidents, poisoning and violence. In total, 206 lung cancer deaths were identified in the cohort. Of these, 191 occurred among individuals who had ever worked underground; the remaining 15 occurred among miners who only worked on the surface. A strong dose-response relationship was observed between cumulative exposure to radon and lung cancer.

The relationship between cumulative WLM and lung cancer risk was modified by time since last exposure, duration of exposure, and dose rate. In contrast, age at first exposure was not a statistically significant determinant of lung cancer risk. After 35 years since the time of last exposure, lung cancer mortality rates among exposed miners dropped to levels experienced by those who worked exclusively on the surfaces.

Twenty-eight (28) lung cancer deaths occurred among men who started working after 1960 (when ventilation was introduced). There was no significant variation in the excess relative risk (ERR) per WLM of exposure between those who started work before 1960 and those who started after 1960. However, the evaluation of cancer risk among those who first started mining after 1960 is based on younger men. As a result, few cancer cases were identified in this subgroup, and therefore, there was limited statistical power to detect an association in these workers.

Morphology data were made available from the Newfoundland Cancer Registry for 97 of the 206 lung cancer deaths that occurred in Newfoundland. Eighty-eight (88) of these 97 deaths occurred among those who had ever worked underground. The histology of these deaths included: squamous cell carcinomas (n=28), adenocarcinomas (n=8), small cell carcinoma (n=7), and other/unspecified (n=45). The small number of outcomes for each histological type did not allow for a reasonable estimation of the ERR/WLM for each type of lung cancer.

Some smoking data were available for 1,107 of the 2,070 miners (53 percent). There was no statistically significant difference in the ERR/WLM between ever and never cigarette smokers. However, stronger associations between cumulative radon exposure and lung cancer risk were noted among individuals who smoked an increased number of cigarettes daily (p<0.05). Specifically, the ERR/WLM was 0.31 percent, 0.46 percent and 0.94 percent among individuals who reported smoking <15, 15-<30 and more than 30 cigarettes per day, respectively. An evaluation of the joint effect of radon and smoking (as measured by cigarettes smoked daily) was unable to adequately discriminate between additive and multiplicative models. However, the data were suggestive of an intermediate relationship (between additive and multiplicative). The evaluation of the joint effects of smoking status and radon was severely limited the small number of lung cancer deaths that occurred among miners who never smoked (n=8).

We also explored the relationship between coronary heart disease (CHD) mortality and radon exposure using both external and internal cohort comparisons. Within the fluorspar cohort, there was no association between cumulative WLM and CHD mortality. A reduced mortality rate of CHD was observed relative to Newfoundland males (SMR=0.86; 95 percent CI=0.74-0.98); this reduction may be attributed to a healthy worker effect.

The assessment of lung cancer risk in this cohort is limited by the relatively small size of the cohort, and hence, a small number of lung cancer deaths. As a result, the standard errors associated with the risk estimates, particularly in multivariable models can be quite large. While smoking data were collected from several surveys, these data were available for only 53 percent of the miners. Moreover, there is undoubtedly some misclassification of smoking status during the study interval as smoking behaviours can frequently change, and the information collected from the surveys did not allow for these changes to be adequately characterized.

There are several important positive features of this cohort. Mortality ascertainment is thought to be near complete and not correlated with exposure status given the ability to link personal identifying information from the occupational files to the Canadian Vital Statistics database. Moreover, the mining population is relatively stable as most miners were born and lived near St. Lawrence for an extended period of time. This improves our ability to identify deaths in the record linkage, and also, to recruit individuals to participate in smoking surveys. Finally, because the source of radon was from water running through the mine, the effects of gamma radiation, thoron and radioactive dust, common exposures in other miner studies, can be ruled out.

Date modified: