Click for next page ( 154


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 153
APPENDIX D THE SUBSISTENCE-FARMER CRITICAL GROUP In Chapter 2 we recommenct that the form ofthe stanciard be a limit to the risk to the average individual in a future critical group. This appendix summarizes the steps that could be involves] in assessing compliance with such a standard! for a particular exposure scenario that defines the critical group as inclucling a subsistence farmer exposed to a maximum concentrator of raclionuclicles in ground water. The risk involved here is the risk of ill health from a radiation dose. Risk entails probabilities as well as consequences. A risk analysis must entail the development of probabilistic distributions of doses to fixture individuals for various times in the future and the development of probabilistic distributions of consequences (health effects) from those dosed. There are various means of constructing risk measures from such probabilistic distributions to be comparer! with a risk limit. The risk measure recommended in Chapter 2 is the expected value of the consequences, determined by integrating the probabilistic distribution of consequences over the entire range of estimated consequences. The conceptual approach to analyzing risks to future individuals from a geologic repository will be illustrateci here for undisturbed performance (e.g., not including human intrusion, meteoric impact, etc.~. Radionuclicles can be released via air or water pathways. The steps in calculating risks for the water pathways are summarizer! here. Similar steps are involves! in calculating risk to fixture individuals via air pathways. For this illustration, radionuclides in waste solids are calculated to eventually dissolve in water and undergo hydrogeologic transport to the saturated zone and subsequently transport via an aquifer to the biosphere. A plume of contaminated ground water will spread! out underground, downstream from Yucca Mountain, to places where it might be susceptible to human use. Calculating the space- and time-dependent probabilistic A probabilistic distribution of a variable can be thought of as the probability per unit increment of that variable as a function of that variable. 153

OCR for page 153
154 YUCCA MOUNTAIN STANDARDS distributions of concentrations of radionuclides in the ground-water plume is the purpose of geosphere performance analysis. Calculation of Geosphere Performance As clescribed in Chapter 3, there are many different possible mechanisms ant! pathways for the dissolution-transport processes. For example, dissolved radionuclides might be transported to the lower aquifer by slow processes that provide time for local sorptive equilibrium with the rock. In other locations, radionuclides might be transported via fast pathways resulting from episodic local saturation, with little time for diffusion into the surrounding rock matrix. The analysis must begin with what might be, in principle, a time- dependent statistical distribution of such scenarios of release and transport. Enough scenarios must be identified that will reasonably sample the events that can contribute to important releases of radionuclides. The probability of each ofthese geosphere scenarios must be estimated so that the resulting analysis can reasonably approximate the statistical distribution of consequences that would be expected. For each geosphere scenario there are large uncertainties in the parameters used in the equations for release ant} transport. For full probabilistic analysis, a state-of-knowledge distribution for each parameter must be developed. Using the equations of transport, these probabilistic distributions of input quantities can be projected into a probabilistic distribution of ground-water concentration, which will vary with position ant! time. Although many useful calculations are made with analytic techniques (NRC, 1983), detailed results require discretizing input quantities, followed by event-tree transport calculations of a large number of combinations of input quantities (EPRI, 1994) or by Monte CarIo/Latin Hypercube sampling of a smaller number of data combinations, as used by the WIPP anti Yucca Mountain Projects (Wilson et al., ~ 994~. Semianalytical adjoins techniques that help create probabilistic distributions from the discretized results are also available. Any of these numerical techniques can yield useful probabilistic distributions, if done properly. The choice is better left to the analyst, who must consider limitations of time, budget, and computer power. Estimates of errors

OCR for page 153
APPENDIX D -THE SUBSISTENCE-FARMER CRITICAL GROUP 155 introduced by sampling techniques shouIc! be incluclec] when such techniques are used to reduce the number of discrete calculations. These space- and time-dependent probabilistic distributions of concentrations in ground water, with emphasis on ground water beyond the repository footprint, are the input quantities needed for calculating radiation doses, consequences, and risks for the biosphere scenarios. Similar approaches are followed for calculating the space and time dependent concentrations of raciionuclides released to the atmosphere. Many analysts employ system software that feeds geosphere results (Erectly into biosphere calculations, bypassing the display of probabilistic distributions of concentrations in ground water. Calculation of Biosphere Performance For the biosphere scenario involving the subsistence-farmer critical group, ground water is assumed to be withdrawn at the location of temporal-maximum concentration of radionuciides. The time of that maximum concentration specifies the time at which the doses, consequences, and risk are being calculated at that location. In the era of temporal-maximum concentration, the concentrations at a given location vary little over a human lifetime, so the ground-water concentration can be assumed constant in calculating lifetime closes and risks for that critical group. The critical assumption in this model, then, is that a subsistence farmer extracts water from the location of maximum concentration of radionuclides in the aquifer, provided that no natural geologic feature precludes drilling for water at that location. The subsistence farmer is assumed to use the extracted contaminated water to grow his food and for all his potable water. Conservatively, the farmer is to receive no food from other sources. A pumped well to extract ground water can perturb the local flow of ground water, so that concentrations of contaminants in the extracted water can be less than in the unperturbed ground water. The extent of concentration reduction depends on the extraction rate (Charles and Smith, 19911. A reasonable extraction rate can be calculated assuming that the subsistence farmer or even the entire critical group uses a single well for extracting ground water.

OCR for page 153
156 YUCCA MOUNTAIN STANDARDS If the subsistence farmer's water is obtained from commercial pumping of the underground aquifer at the point of maximum local contaminations, the effect of commercial rates of water extraction on the withdrawn concentration can be included in the analysis. Obviously, for commercial water withdrawal, it is the withcirawal location rather than the location of the subsistence farmer that is important. The vertical variation of concentration in grounc! water at a given surface position can be obtained from the geosphere analysis. If methods of predicting the vertical location of the point of water withdrawal within the aquifer are clefensible for the long-term future, then the effect of withdrawing at locations other than that of the vertical maximum concentration can be included. Otherwise, arbitrary assumptions of well depth would diminish confidence in the resulting calculated risk. The largest radiation exposure to fixture humans from contaminants in grounc! water is predicted to result from internal radiation from ingested or inhaler! radionuclides. For the water pathways, eating food contaminated by irrigation or by other use of contaminated ground water for growing food is expected to be the source of largest dose, greater than closes from drinking water (NRC, 1983~. Therefore, realistic prediction of closes and risks to future humans requires knowlecige of their diets and amounts of fooci and water consumed. Such information for the distant future is unknowable. Therefore, as is clone in all other biosphere scenarios, we must assume that future humans have the same diets as ourselves (including foot] and water consumption). This amounts to the unavoidable policy decision that geologic clisposal is to protect future humans whose diets are the same as ours or whose diets would! not lead to greater radiation closes from using contaminated water than would the filets of people today. All biosphere scenarios must also rely on tiara for the uptake of radionuclides from contaminated water into food. Here, one can rely on scientific data for the typical soil conditions and for the kinds of foods assumed for this analysis. For a given food chain and for drinking, the amount of radioactivity ingested in a given time, or over a human lifetime, 2 There is a current proposal for commercial withdrawal of ground water Tom the aquifer near Yucca Mountain. This water could be distributed to local communities as well as others that might exist or be developed farther from Yucca Mountain.

OCR for page 153
APPENDIX D - THE SUBSISTENCE-FARMER CRITICAL GRO UP I 5 7 is proportional to the concentration of radionuclicles in the extracted ground water.3 The ingredients of the biosphere approach describer} here, beginning with specified concentrations in extracted ground water, are iclentical with those of the widely used GENI computer code cleveloped by Napier et al. (19881. The GENT code is used by the WIPP Project in predicting closes to future inclividuals who utilize contaminates! water for drinking and for growing foot] ant} who receive no food from outside sources. It is an example of what could be used or updater} for calculating subsistence-farmer doses. The GEN! code includes intake-close parameters recommender} by ICRP and other agencies. Therefore, employing GENI or a similar code to predict radiation doses to future humans who inadvertently use contaminated water requires the adclitional assumption that future humans have the same dose-response to ingested radioactivity as clo present humans. All biosphere scenarios adopt this assumption. Of course, it is expected that the intake-dose parameters will be updated when new information is available. Given the probabilistic distribution of concentration of radionuclides in extracted grounc} water at a given future time and location, the human-uptake-response mociel, such as GENI, can predict the statistical distribution of radiation doses to the subsistence farmer. Because the grounci-water concentrations vary little over a human lifetime, it is necessary only to sum the close commitments for a human who uses that contaminated water over his/her lifetime. The result is a probabilistic distribution of lifetime dose commitments, easily converted to lifetime average annual close commitments. The probabilistic distribution of lifetime dose commitments can be converted into a distribution of consequences by multiplying each value of dose commitment by the appropriate dose-risk parameters, obtainable from ICRP and others. If the constant dose-risk parameter of the linear hypothesis is used, the probabilistic distribution of consequences will differ from that of doses by only a constant multiplier. Here, by adopting dose 3 This assumes uptake factors, i.e., distribution coefficients for a given radiochemical species in a given plant or other organism immersed in contaminated water, that are independent of radionuclide concentration.

OCR for page 153
158 YUCCA MOUNTAIN STANDARDS risk parameters developed for present humans, we are assuming that fixture humans will have the same present risk when exposed to a given radiation dose. All biosphere scenarios adopt this assumption. Of course, it is expected that the close-risk parameters will be updater] when new information is available. Each value ofthe consequence is then multiplied by the probability distribution function for that consequence, and this integrand is then integrated over all consequences. The result is the calculated risk to the subsistence farmer from ground-water pathways, expressed either as the lifetime risk or as the lifetime average annual risk. To this risk from the ground-water pathways are to be added other calculated risks for the subsistence farmer, who is the individual at maximum risk within the critical group. To obtain the risk to the average member of the critical group, for compliance determination, it can be arbitrarily assumed for simplicity that there is a uniform distribution of inclividual risk within that group.4 Because {CRP's homogeneity criterion specifies that the critical group should have no more than a tenfold variation in inclividual dose, and because large departures from the linear dose-response theory are not expected for this calculation, the expected value of the risk to the average individual will be about ore-half that ofthe maximally exposed subsistence farmer.S The expected value of risk to the average individual within the subsistence-farmer critical group is to be compared with the risk limit that is to be selected for compliance. The regulator can specify how far below 4 Adopting any distribution, uniform or otherwise, for the risks within a critical group projected to exist in the distant future, cat 100,000 years and beyond, is arbitrary, because the habits, location, etc. of that future group of people are not knowable to us. Whether one postulates some distribution, as is done here, or calculates a distribution based on the assumed relevance of the current site- specific population, adopting any such distribution for the future is arbitrary. 5 Because of the large uncertainties in the calculated doses and risks to any of these individuals, the uncertainty of uniformity of risk within the group cannot introduce an important uncertainty in the result. An uncertainty of 2 or 3 in the calculated dose is not expected to be important.

OCR for page 153
APPENDIX D -THE SUBSISTENCE-FARMER CRITICAL GROUP 159 or above the specified risk limit the calculated risk must be for compliance clecision.6 6 UK's NRPB specifies the calculation of a 95/0 confidence interval for the expected or central value of risk. The upper value of this confidence interval is what is compared with a regulatory limit [Barraclough et al., 19921.

OCR for page 153