The radionuclides released from an underground nuclear explosion are initially distributed by direct explosive action in the immediate vicinity of the explosion. At some time, these nuclides may be transported by ground water to sufficient distances to raise problems of water management. Reviewed separately are: (1) the nature and amount of radionuclides produced by underground nuclear explosions, (2) the initial distribution of the biologically significant nuclides and (3) the possible transport of these nuclides by ground water. For purposes of this discussion it is assumed that a 50-kiloton part-fission part-fusion explosive will be used.

Activation Products

The exact amount of each activation product generated by the neutron flux from a nuclear explosion depends

on the chemical composition of the rock. The activities induced in average crustal material are relatively shortlived. (32.33) At the end of one year, the induced activities, except for Co, are insignificant.

The composition of oil shale from a drill hole near the depositional center of the Piceance Creek Basin is given in Table 8. It is probable that the composition of the shale anywhere near the center of the Basin would be similar. In general, with increasing distance from the depositional center, the amounts of nahcolite and daw*Appendix A is a general review of the ground water system of the Piceance Creek Basin as it might be affected by an underground nuclear explosion, and was prepared at the request of the San Francisco Operations Office of the AEC. A study of the problems directly related to Bronco would be conducted by the Nevada Operations Office of the AEC when the technical concept and the shot location have been established.

sonite would decrease, and would be replaced by increasing amounts of quartz and clay minerals. Although the saline-rich zone in the Piceance Creek Basin contains a relatively high concentration of sodium, at 4 to 5 percent (by weight), it is not significantly greater than the average amount of sodium in the earth's crustal rocks at roughly 3 percent. The activation product, Na, with a 15-hour half-life, would not be present in significant amounts after the explosion and within a few months, it would be difficult to find a trace of Na24. The activation product, Fe with a 45-day half-life, would not be present in significant amounts as the Fe 0, content of the oil shale is not more than about 3 percent, approximately the crustal abundance.

The activation product, Co, with a 5.2-year half-life, would be present in minor amounts, as oil shale probably contains 1 to 2 parts per million of cobalt. The total amount of Co from a nominal 50-kiloton nuclear explosion might be on the order of 100 curies.

The small amount of nitrogen in the oil shale, less than 0.5 percent, would lead to a trivial amount of C11, insufficient to warrant consideration.

Fission and Fusion Products

At the end of one year Sr and Cs137, both with about 30-year half-lives, are the principal remaining fission products of recognized biological importance. It is assumed that Sr is the more significant radionuclide, as Cs' is more firmly held on the solid by exchange mechanisms than is Sr. The gaseous fission products, Xe133 and Kr, are not considered significant as potential contaminants in ground water. If one-half of a 50-kiloton explosion were fusion energy release, about 25 × 101 curies of tritium with a half life of 12.3 years would be produced.

Initial Distribution

The initial distribution in oil shale of 4 x 103 curies of Sfrom the 25-kiloton fission energy release can be calculated on the basis of four conservative assumptions: (1) the radius of the cavity is 100 feet, (2) the radius of the fractured zone is 220 feet, (3) the density of the oil shale around the point of explosion is 2.2 (g/cc), and its porosity (water saturated) is 0.02 (2 percent), and (4) radionuclides are distributed only in the chimney by direct explosive action, and post-shot collapse of the chimney into the cavity will not affect the nuclide distribution.

On these assumptions, the mass of solids in the chimney will be 2.38 x 10 grams, the mass of water in the

pore space will be 2.2 x 1010 g. (or ml), and the total mass will be 2.4 × 1012 g.

Assuming that Sr9 is all soluble and uniformly distributed throughout the oil shale in the chimney, its initial concentration in the total mass would be 4.0 × 103 curies in 2.4 x 1012 g, or 1.7 x 10-9 c/g.

When equilibrium is reached in the exchange of Sr90 between the oil shale matrix and the contained pore water, the amount of Sr in the water is expressed by the equation:

where Ka, the distribution coefficient for Sr in oil shale, is estimated at about 100. The Sr9 activity in the water then 0.37 curies; and the Sr0 concentration in the water, 0.37 curies in 2.2 x 1010 ml, is then 1.7 × 10-11 c/ml.

The only long-lived activity of possible importance induced in oil shale would be Co (5.2-year half-life) at no more than 100 curies. Thus, 100 curies would be distributed throughout the crushed zone, and its initial concentration would be 4.5 x 10-11 c/g. Co is analogous to Sr in exchange behavior, and after reaching exchange equilibrium between oil shale and the pore water, its concentration would be 4.5 x 10-13 c/ml.

The fusion reaction in the explosion would produce 25 x 10' curies of tritium (H3), the preponderance of which would be in the form of tritiated water. The H3 concentration in the pore water would be 25 × 101 curies H3 in 2 × 1010 ml, or 1.25 × 10-3 c/ml. It is assumed conservatively that this tritium concentration will be reduced by radioactive decay and by dilution with ground water outside the chimney, but not by exchange mechanisms.

Transport of Radionuclides

Evaluation of the transport of radionuclides by groundwater solutions requires: (1) determination of the velocity and direction of regional ground water flow, (2) determination of chemical composition of the ground water and of the physical and chemical properties of the rock matrix, and (3) determination of the specific Ka's for specific radionuclides, using representative samples of the rock and the contained ground water.

For the Piceance Creek Basin, the velocity and direction of flow are not well known. A reasonable assumption is that the flow rate is from 10 to 100 feet per year, radially inward from the structural rim and then northward via Yellow and Piceance Creeks drainage. Chemi

cal analyses of representative ground water from the oil shale indicate a high sodium bicarbonate type with minor amounts of Ca, Mg, and K. (34)

The distribution coefficient, K., is defined as the ratio of the concentration of a particular radionuclide on the solid to the concentration of that nuclide in the adjacent solution, at equilibrium. As ground-water solutions move away from the saline-rich and oil-rich zone in the center of the basin, where presumably the nuclear explosion will occur, the clay mineral content will markedly increase with corresponding decrease in the carbonate minerals. It follows that the Ka for Sr90, estimated to be about 100 if the point of explosion is in the high saline zone, will tend to become larger, probably in the range of 500 as the clay minerals become abundant. For the proposed 50-kiloton explosion in the oil shale the transport of the radionuclides Sr and H3 can be calculated to a first approximation, assuming that: (1) the ground-water flow rate is 100 feet per year, (2) the value of Ka is 500, and (3) the flow is completely laminar. The average flow rate of a single nuclide such as Sr is related to the flow rate of the ground water by the equation (35)

it would not enter appreciably into ion exchange reac tions, although as tritiated water it might exchange with chemically bound water in the rock matrix, and thereby be slightly retarded in respect to the ground-water flow rate, by a few percent in clean sands to possibly 50% in a rock high in clay minerals.

It is reasonable to assume that the travel path of ground water from around the point of explosion to where it contributes to surface water flow would be on the order of several miles - for convenience, assume 5 miles or 25,000 feet. At a ground-water flow rate of 100 feet per year, travel time to the outlet would be roughly 250 years. For Sr in the ground water, an additional retardation factor of 50,000 must be used; obviously by the end of this time period, the Sr0 would have completely decayed. For H3, with a 12.3-year half-life, the 250-year period would lead to a decay to 1 x 10-6 of the original activity, or a reduction of six orders of magnitude.

Conclusions

1. On the basis of the available data and assumptions made, no contamination of surface water sources seems probable.

2. Should it become necessary to remove water from a rubble-filled chimney, a water disposal problem may arise, due to the Sr0 and H3 content in the initial water flow from the chimney. Therefore, it is recommended that the area surrounding ground zero be dewatered prior to detonation.

It is known from experience with many underground nuclear tests that most of the fission products and induced radionuclides produced during the nuclear blast will be trapped in the fused rock (puddle glass) that accumulates at the bottom of the chimney. The crushed shale, however, will be contaminated with fusion produced tritium, presumably mostly as tritiated water and certain fission products, for example Sr and Cs137, which have gaseous precursors, and Ru106 which forms volatile compounds. The quantity of these radionuclides which may appear in the product oil and in the gases which are produced during the retorting are important to the future of nuclear explosives for crushing shale. This appendix presents the initial results of laboratory tests which are currently being carried out by ORNL to give an indication of the fate of the radionuclides during shale retorting. However, without an actual nuclear test it is impossible to assess in detail the potential problems of industrial safety involved in producing and handling the oil.

QUANTITIES OF RADIONUCLIDES PRESENT. The amounts of the individual fission products and tritium produced by a detonation of a given yield depend on the type of explosive used. This evaluation assumes that most of the energy would be derived from fusion. If retorting starts about 15 months after the shot, the tritium activity will be more than 95% of the total activities present. In addition, a small amount of radioactive material formed by neutron activation of the shale surrounding the explosives would be present. Irradiation of a sample of Green River oil shale in the Oak Ridge Research Reactor indicated that Sc46, Fe59, Mn, and Zn would probably be the most important of the long-lived induced radionuclides. As indicated above it is anticipated that most of the tritium (as water) and significant quantities of Ru106, Cs137, Sr90, and Y91 will be present in the shale rubble. The gas within the void volume will contain tritiated water vapor (a small fraction of the total tritiated water), tritium gas and Kr. Most of the remaining fission and activation products should be trapped in the puddle glass.

CONTAMINATION OF THE RETORT-OFFGASES. The gases initially emerging from the retort during retorting will include krypton, but more significantly, tritium, assumed to be mostly as tritiated water vapor. It is estimated that a 50 kt shot would produce large quantities of tritium and, because its half-life is 12.3 years, this amount would nearly all be present when retorting starts. The concentration of radionuclides in the gases and consequently the handling of the gases depends on a number of factors. For example, essentially all of the krypton and that portion of the tritium present as tritium gas should be flushed readily from the chimney. The rate at which the tritiated water will be removed will depend upon the extent of its diffusion into the shale and subsequent equilibration with the water bound in the shale. It is possible that a significant portion of the tritiated water will remain at the shale surtace in a form which can be evaporated and can thus be removed by passing a relatively few void volumes of hot gas through the chimney. Tritiated water bound in the shale, however, will be released slowly; initial laboratory tests indicate that, in retorting, the bound water is released at a temperature only slightly lower than that of the oil. In addition, the tritium concentration in the gas stream will be affected by the ratio of the volume of gas recycled to the retort to the volume discarded. In any case, in designing the facility, careful consideration must be given to providing for dilution of the offgases to an acceptable level prior to discard or for dispersing them to the atmosphere in a manner which will ensure adequate dilution prior to contact with on-site personnel or with the off-site population.

CONTAMINATION OF THE OIL. Conceivably, the product shale oil could be contaminated with radionuclides at several points in the process, for example:

1. Exchange of tritium and hydrogen between water and the shale oil during the retorting phase.

2. Dissolution of oil-soluble fission product compounds present in the shale rubble and puddle glass.