levels of radioactivity in a room where such natural gas would be bumed is predicted to be about at the RPG levels permissible for continuous exposure to the general population. However, by flushing the chimney prior to introducing storage gas, these levels can be reduced several orders of magnitude, and would thus be far below the Radiation Protection Guides developed by the Federal Radiation Council for control of normal peacetime operations. The use of dehydrators and particulate filters at the wellhead can be utilized to remove any refractory radionuclides and tritiated water entrained in the gas.

A nuclear gas storage experiment would be executed under the control of the AEC which would retain direct responsibility in matters of public safety, including radiation. The AEC has not developed regulatory limits which are directly applicable to the gas storage application and it is expected that the results of the ex

periment would be used as a partial basis for developing such limits. These limits may be some small fraction of the FRC guides or of the recommendations of the KRP and NCRP. After satisfying the experimental requirements, any commercial use of storage gas from the chimney containing radioactivity would be subject to appropriate regulatory approval. Such approval would be granted only after a determination has been made that use of the gas would not result in a significant increase in the radiation exposure normally received by the general public.

Fuller discussions concerning removal of residual solid and liquid radionuclides by filtration, flushing of gaseous radionuclides from the chimney, exchange of gaseous radionuclides with atoms in the storage gas molecules, and other radioactivity considerations may be found in the "Technical Program" section of Chapter VI and the "Safety Evaluation" section of Appendix D.

The usefulness of a gas storage reservoir depends on its ability to meet peak short-term demand situations and/or its ability to satisfy requirements over an extended period of time. The former is generally referred to as its peak deliverability; the latter as its annual turnover capacity. Storage reservoirs are always operated in conjunction with a given level of contracted supply which originates in a pipeline or a producing area. given type of storage system may be economically attractive in satisfying deliverability or turnover. It need not satisfy both requirements--particularly if it is operated with some other facility that can satisfy the deficiency in the other sector.

A

The attractiveness of developing storage depends on the pipeline load factor under which gas is supplied and the cost of storage. If gas is sold at 100 percent load factor (at maximum capacity every day, all year long), there is no economic justification for storing gas at all except as an emergency supply. On the other hand, if the load factor is 50 percent, i.e., the annual amount sold is 50 percent of what the pipeline is capable of delivering, then there is considerable incentive to develop storage. The value of identical storage facilities may

be substantially different for companies selling gas at different load factors.

To demonstrate the economic position which nuclear reservoirs might assume, the cost of nuclear storage has been developed on a unit basis. That is, the investment cost per Mcf of peak deliverability and of turnover gas has been developed by making some provisional assumptions and this cost compared with that of conventional storage systems. What is more important, however, is what this facility is worth to a given purchaser of pipeline gas. This, too, will be shown by comparing the economic impact on different companies that are purchasing gas at different load factors.

To make a realistic comparison of a nuclear storage reservoir against competing alternatives, it is necessary to estimate costs required to create such a field on a reasonable commercial production type basis. It is also necessary to calculate the cost of completing the nuclear chimney as a gas storage reservoir. This has been done for three sizes of yields in Appendix A using the gas storage capacity figures developed in Table III. These costs are summarized here (Table IV).

It should be noted that in using larger explosives, the costs of creating a reservoir rise much more slowly than the increase in storage capacity. Most of the costs involved are fixed regardless of yield variation. Only charges for nuclear explosives, 13 emplacement costs, and safety vary to any significant degree with yield and these are proportionately small. For example, substitution of a 100-kt explosive for a 24-kt explosive might increase costs for creating nuclear storage 15%-25%. The value of the reservoir created for gas storage purposes in terms of capacity might be enhanced by as much as 400%.

The future development of nuclear storage depends upon what the ultimate production-type project will cost and how limiting the seismic and radioactivity factors will be. Seismic problems and, to some extent, radioactivity problems are a function of the geographic site selected. Their costs are also subject to technological advances.

In general, there is no question that conventional gas storage in depleted gas reservoirs is at present the

most economical method of supplying both peak day deliverability and seasonal turnover requirements. As will be shown, nuclear storage may provide an economic breakthrough, particularly on the basis of deliverability. There are many variables in operating conditions and design parameters that make detailed cost comparisons complex. To make a reasonable comparison, investment costs have been shown for comparable facilities to those conditions used in designing storage facilities for the 24-kt and 100-kt nuclear gas storage fields. Because underground storage facilities do not have the same combination of performance characteristics, comparisons are shown on the basis of allowing turnover and deliverability requirements to become controlling in turn.

In Table V representative investment costs 14 for comparable types of different kinds of storage and supplemental supply are given. They show that low pressure holders, pipe batteries, and bottle storage are far too expensive even to be considered for any large application.

They also indicate that on plant cost basis, both LPG and LNG systems are substantially more expensive than high pressure underground storage in natural reservoirs. Actually, on a plant cost basis, LPG looks somewhat cheaper than LNG. However, in terms of total cost, they are not far apart because in LPG systems the expense of propane and that associated with interchangeability problems add to the total cost. The underground storage field cost estimates shown in Table V are based on the lower range of current costs. Nuclear storage costs are substantially lower in both peak day deliverability and annual turnover costs in the larger yield sizes, and lower in peak day deliverability in the lower yield range. Turnover costs in the lower yield range are slightly higher than conventional storage fields. However, if conventional storage reservoirs are not available, then a nuclear reservoir appears to be a much more preferable alternative, even in the smaller nuclear explosive yield range, than either LNG or LPG. For this reason, as well as because of the increasing scarcity of satisfactory low cost conventional fields, nuclear reservoirs have an obvious application in many areas. In regions without natural reservoirs, the application could have a tremendous impact on the industry. Even in areas which have heretofore had plentiful storage, the impact could be substantial.

At this time, it is not known in detail how to get the optimum advantage from a nuclear gas storage reservoir. It is probable that the additional advantage of the high deliverability and, hence, winter refillability of a nuclear reservoir could generate even greater economical advantages than indicated.

In Appendix B the costs of several recently developed or considered conventional underground storage fields in the Columbia Gas System have been compared to potential nuclear gas storage fields in detail.

The value of a nuclear storage field must obviously be in excess of its cost for it to be considered as a factor in gas supply. The cost comparison figures on the previous page show that nuclear storage may represent an economic breakthrough in solving peak day deliverability problems under any conditions. They also indicate that in most turnover applications (except

in small scale operations in direct competition with conventional underground storage systems) nuclear storage has a promising future. What they do not indicate is the full scope and variation in potential value to the gas industry in different situations. Even if costs significantly below current levels are not achieved, nuclear storage would be attractive in many situations because the alternatives are less satisfactory. The limit of this applicability will, of course, depend on the costs that can be achieved. Higher costs than those indicated could restrict applications to fewer situations. Lower costs might even result in replacing what have been considered up to now as satisfactory alternative supply systems.

To show the variation in value as separate from cost, the ownership of the 24-kiloton test reservoir is evaluated as if it would be owned by a gas transmission company and by a gas distribution company. The detail of this comparison is given in Appendix C for a representative transmission company and for a representative distribution company in Pennsylvania. If the nuclear field were owned and operated by a transmission company with substantial gas storage facilities and a 95% supply load factor, the breakeven value of a 24-kt nuclear reservoir is projected to be $6 million. That is, an investment of less than $6 million would be economically justified. This is substantially in excess of predicted future costs of a 24-kt nuclear reservoir.

If the nuclear field were owned by a gas distribution company with a supply load factor of about 50%, then the investment in a nuclear reservoir that could be justified would be something less than $22 million. However, the ability of a distribution company to take advantage of this situation would depend on the market concentration. More than likely it would require additional and possibly extensive pipeline and compression facilities to utilize the full benefit of the storage facility.

The difference between $6 million and $22 million is substantial and illustrates how the value for different supply situations can be vastly different. A distribution company with a low load factor can afford to spend a considerable amount in supplementary storage and pipeline facilities and still benefit substantially.

A. General

Project Ketch is a planned experiment designed to clarify several technical uncertainties and to determine whether or not it is economically feasible to create underground gas storage with nuclear explosions. The technical problems to be studied are set forth under Section D of this chapter. Evaluation of operations during the final phase of the project will determine economic feasibility as applied to commercial application. Project Ketch, designed to meet these objectives, was developed by personnel of the Lawrence Radiation Laboratories, K Division, with the assistance of the Columbia Gas System Service Corporation and is described in detail in the Technical Concept (Appendix D).

This experiment, if approved, would utilize a 24kiloton explosive to be detonated at a depth of 3,300 feet in a section of rock known as the Chemung Shale Formation. This detonation would be expected to create a roughly cylindrical rubble chimney with a radius of ninety feet and a height of about three hundred feet. Permeable cracks would extend for some further distance above and around this chimney. The maximum extension of such cracks above the shot point would be expected to be no more than 650 feet. The rock medium itself is considered to be of low permeability to gas on all sides (see "Geology of the Test Region,") so that if natural gas were injected into the chimney under high pressure it should be satisfactorily contained within the newly created reservoir.

Present indications are that such a project could be carried out safely. The initial steps of the project would consist of a drilling program and safety studies to confirm site acceptability. Major activity leading to project execution would be contingent on such confirmation. Extensive engineering data have been developed on the physical effects of nuclear explosions in various rock types from over 200 underground detonations in Nevada, New Mexico, Alaska, and Mississippi. The rock types involved have included those of greater and less density, more and less water content, greater and less friability, etc. than the Chemung Shale Formation. From these data code calculations have been developed which can handle any silicate rock, making possible accurate predictions of the effects of a shot in the Chemung Shale. Moreover, in the fall of 1967, a closely related experiment on gas stimulation, Project Gasbuggy 6, is

scheduled to be carried out in New Mexico. Data obtained here relative to radioactivity in the product gas would be particularly applicable.

B. Location

The location chosen for study purposes is in Clinton County in north central Pennsylvania, about 12 miles southwest of the town of Renovo. The site is within the boundaries of the Sproul State Forest, a heavily wooded, sparsely populated region. (See Figures 5,6)

This general region was selected for the study because it is located near a major gas pipeline, which would be necessary for subsequent testing, and is situated in a region near a market area where gas storage is needed. The geologic environment is also favorable and the population density low. These conditions were considered to be representative of other prospective areas which might be used for future nuclear gas storage fields. The specific area selected would, in any case, be provisional, in that a test program requires confirmation of geological factors from detailed exploratory work. If a site does not meet these requirements, a new location would be selected.

C. Geology of the Test Region

The site is located on a large anticlinal structure known as the Hyner Dome 15. (Figure 7) This structure was drilled in the early fifties for natural gas by the Manufacturers Light and Heat Company (an operating company of the Columbia Gas System) and others. While no commercial quantities of gas or oil have ever been found in this immediate area, valuable geologic information has been acquired. Scattered soil that has resulted from weathering of the bedrock covers the surface. The surface rock 16, 17 in this area is white Pocono sandstone of the Mississipian age. This is underlain for a few hundred feet by the Oswayo shale formation of Upper Devonian age.

Beneath this lie almost six thousand feet of other Upper Devonian sediments, predominantly red and brown shales and slaty shales. At the bottom of this section there is about 100 feet of well-defined hard grey Tully limestone. This is followed by about one thousand feet of Middle Devonian shale and sandstone culminating in the Onondaga limestone. Just below this lies the Lower Devonian Oriskany sandstone. The top of the Oriskany lies at about 8,285 feet beneath the surface of the test