ML18192A432

Regarding Report, Palo Verde Economics-APS Projections Versus National Averages, Further Discusses Construction Cost, Fuel Efficiency, Fuel Availability, Capacity Factor and Decommission Costs

ML18192A432
Person / Time
Site:	Palo Verde
Issue date:	06/30/1977
From:	Lowes R Arizonans for Safe Energy
To:	Dircks W NRC/EDO/AO
References
Download: ML18192A432 (79)

Text

FROtP.WSCekf 0>>Lycee Ac'&mane'fee Safe-Zee~P~eiX, At'kl TO: CD,X~Z.Qkxcke ACTION CONTROL DATES COMPL DEADLINE ACKNOWLEDGMENT

'NTERIM REPLY FINAL REPLY FILE LOCATION/.I s/P i'7 CONTROL NO.DATE OF DOCUMENT PREPARE FOR SIGNATURE OF: Q CHAIRMAN Q EXECUTIVE DIRECTOR OTHER: DESCRIPTION

~LETTER Q MEMO Q REPORT Q OTHER 544pscscs Bxcak+4>05tch BRC,%53+'ccQ QK bf9 Xe~ie aCI the Wtz<@4 Ce"ya~~htee CCe=ieSXOa X6.@@SR@CIf.4'ba PRO Ve"de pXeec SPECIAL INSTRUCTIONS OR REMARKS DOCUMENT/COPY NO.NUMBER OF PAGES POSTAL REGISTRY NO.CLASSIFIED DATA C LASS I F I CAT I 0 N CATEGORY Q NSI Q RD Q FRD ASSIGNED TO: DATE INFORMATION ROUTING LEGAL REVIEW Q FINAL 0 COPY ASSIGNED TO: DATF NO LEGAL OBJECTIONS NOTIFY: Q EOO ADMIN8c CORRES BR EXT."'OMMENTS, NOTIFY: EX j.JCAE NOTIFICATION RECOMMENDED:

Q YES 0 NO NRC FORM 232 (11-75)EXECUTIVE DIRECTOR FOR OPERATIONS PRINCIPAL CORRESPONDENCE CONTROL DO NOT REAIOI/E THIS COPY

iX William J~Dircks, Assistant Executive Director for Operations U.S.Nuclear Regulatory Commission ROOM m.6.~ETU>>TOREGULAT aV CzNTzm ZrrZ"~e r~~~~S~~In late, April of this year, I had sent a copy of my report (Palo Verde Economics-APS Projections Ver'sus National Averages, 4 Aprilj77, Encl.>I)to my representative, Mr.John Rhodes, which he kindly forwarded to the NRC.I received a copy of the reply you sent to Mr.Rhodes (Docket j" 50-528-30)

May 25, l977.The reply contained many points that I wish to comment upon at this time.In this letter, I will discuss these areas which we disagree on: (a)construction cost;(b)fuel efficiency;(c)fuel availability;(d)capacity fa'ctor;and (e)decommissioning costs.In your letter your office stated that"It is in the staff's view, implausible to'add further escalation of$1.6 to$6.2 billion as Mr.Lowes does." I am assuming that your staff projects a final construc-tion cost of$2.9 billion with a maximum cost escalation cost of$l billion, thus giving a base cost of$I.9 billion.Given this estimate, then this$I.9 figure is less than what the plant would have cost had it been finished in 1975.I must point out that at the$2.9 billion cost, the per kWe (in-stalled)cost woul d be$761.15 per kWe.The Federal Energy Administration states the cost of building a plant would be$640 per kWe in l975 dollars.(Robert I.Hanf ling, Deputy Assistant Administrator, Energy Resource Department of the F.E.A., Washington, D.C.;taken from letter sent to Rep.Eldon Rudd, 7 June 77).At$76I.I5/kWe, the escalation rate would be approximately 2%.I should note here that I made a miscalculation in my report.I reported that Edward Cowan of the New York Times stated that the l976 cost of building a reactor was$773/kWe.I have since discovered that Mr.Cowan meant the$773 figure as the average cost projection.

He also stated that the average cost projection in 1967 was$134.So, usSing tsat$134 Figure for'67 and the$773 figure for'76, it can be seen that the average projected cost escalation over that period ('67-'76)is approximately 2I.5%.Thus, I find it unacceptable to assume over the next eight years a mere 2%annual escalation rate.In my report I listed the reasonable annual escalation rate as 5%to l5%.That'rojection:should stand--or both figures should be raised.As an additional note, I received a reply to my report from the F.E.A., quoting a$I245/kWe construction cost figure for a plant in New York.(Completion date I985.)Applying that figure to Palo Verde, I come up with a construction cost of$4.45 billion completed (calculation includes a,6.9%deduction to bring it down to 1984--6.9%

is the F.E.A.'s calculated escalation rate.)js, F r L S.'I I f'L j//I IL I L L To William J.Dircks Page 2 The NRC, AEC and utility companies, have been notorious for making economic estimates that unrealistically post a benefit to nuclear costs, in the environmental impact statements released for the now-completed plants and even the nuclear plants now being built.In fact, a study by the Massachusetts Institute of Technology states that the average total cost escalation For nuclear plants from the original projected price to the final construction cost has been 10(P/c.The Koshkonong plant in Wisconsin and the Consumers Dow plant in Michigan have both in-creased in projected price by over I00/o.The Wisconsin plant has not even been started, while the Michigan plant is still in the construction stage.The NRC's Office of Operations did not respond to the subject of fuel efficiency in the reply, but I would like to turn your attention to it since I have recently uncovered some new findings.I stated in my report that the average fuel efficiency is l4 million kilowatt hours of electricity (MMWhe)per short ton of milled uranium (yellowcake).

These points will support my conclusion that 14 MMWhe per short ton of milled uranium (yellowcake) is the average fuel efficiency ratio.F.B.Baranowski, Director of the Division of Production and Materials Management of the former Atomic Energy Commission, says that for l97I through l973g the ratio was 14 MMWhe per short ton of yellowcake (M.C.Day, Bulletin of the Atomic Scientists, Dec;'75, p.53).Also, according to Morgan Huntington, Director of the U.S.Bureau of Mines, the average ratio is about l4 MMWhe.Mr.Huntington states that the maximum to be expected should be 22'hich is significantly lower than the approximate 25 MMWhe per ton of uranium used by the NRC in the Palo Verde Final En-vironmenta I Sta tement.A report by Ron Carstens and Robert Lamson (Oct.'76, Box 37, Anacortes, Wash.9822I)titled Realistic Uranium Energy Yields and Cost (Encl.<<2)l" conce'r'ris)energy yields of reactors already in operation.

The team requested information on uranium requirements from companies with reac-tors five years and older.Seven companies replied.Carstens and Lamson concluded that the av-erage energy yield from yellowcake is l2.36 MMWhe (weighed average by length of operation).'he report,"Uranium Reserves, Resources and Production" (June 15,'76)cited by your office states that the amount of uranium ore available in the United States is enough to fuel three hund-red l,000-megawatt reactors over their entire lifetimes.

Since you state also that there are I.84 million tons of reserves, it can be deduced that by divid-ing the amount of electricity that the units would produce (at the NRC projected 75%capacity factor for the Palo Verde)by the amount of known and probable reserves (I.84 million tons)yields an energy efficiency ratio of 32 MMWhe per short ton of milled uranium.Using a more reason-able capacity factor of 40%for the average I,000-megawatt reactor, the fuel shortage cannot be alleviated but only relieved temporarily.

'lso the report cited by your office states that there is enough uranium to fuel any reactors which may be placed in service by l990.This statement does not address the time-span used in my report.Using a 35-year life span for the Palo Verde plant, the reactor units will operate up to 20I7 to 202I.This means the Palo Verde will be operating 27 to 3I years after the l990 date.James Schlesinger of the Carter administration announced that 550 nuclear plants would be on line by the year 2000.According to the Chicago Tribune (April 24,'77),"Schlesinger proposed con-struction of 200-300"new nuclear plants in the next fifteen to 20 years."

~~4' To William J.Dircks Page 3 As for the reliability, of previous federal government estimates on fuel reserves, the expected costs for low-grade uranium hav'e gone up,from$50/lb.of yellowcake at 60-80 parts per million (see Encl.>3)to$I00/lb.at 100 ppm in a two-year period.(John Kiemenic,"An Estimate of the Economics of Uranium Concentrate from Low Giade Sources", Monograph, Planning and Analysis Divisio'n,'Grand Junction office', USACE, Grand Junction, Colo., Oct.22,'74).'I Concerning my"projected price of fuel for the Palo Verde, Carstens and Lamson project a fuel cost of 20 mills per kWhe in l985--neglecting any possibility of a fuel shortage.My report projects a somewhat severe fuel shortage and assumes a minimum cost of 28.57 mills per kWhe.Saunders Miller, in a just-released comprehensive study of nuclear and coal power economics, concludes"Unless more'known reserves're found, some of these reactors may have to shut down before their economic lives are completed.

Prudence would dictate that no more new reactors be started until'potential resources'ecome additional

'known reserves'".

The book,"The Economics of Nuclear and Coal Power" by Miller, was reviewed by Baron's May 30, l977.(Enclosure

<4.)The review author calls this study very reliable and"very conservative" in estimates.

Saunders states that the amount of l000 MWe reactors that could be fueled from the"known","probable","potential","possible", and"speculative" reserves would be 624 for their 30-year lifetimes.

It is important to note that Saunders assumes a 65~/o capacity factor and a fuel effi-ciency ratio of approximately 30 million MMWhe per short ton of yellowcake.

If the 65~/o capacity factor is brought down to 40/o and the fuel efficiency is brought down to l4 MMWhe per short ton, the number of reactors is changed to 464.It must be emphasized that the"'potential resources't this time-'probable', possible', and'speculative' are only'maybe's'", Saunders rei tera tes.Concerning Capacity Factor, by negating the poor performance of the Brown's Ferry reactors in the total sample of reactors over I000 MWe in size, your office conveniently deduces that the presently existing reactors are performing much better than they, in a complete sample, actually are.It has been suggested by both industry and government agencies that the poorly performing reactors such as the Brown's Ferry and Palisades plants, be dropped from the sample when pro-jecting capacity factors of future nuclear units.The following are reasons why exclusion of such reactors would constitute a poor sample: I)Onewf-a type incidents are consistently occurring that decrease the performance of individual reactors.2)There still, after years of commercial reactor experience, is no large-scale.

stan-dardized design for reactors presently" being built or reactors planned for the future.The plants that are now performing poorly are supposed to have been better than their predecessors

-just as the reactors being built now are supposed to be better than their predecessors.

In this view, the expected capacity factors for the reactors under construction should not be any better than the capacity factors of the reactors already in operation.

3)There has been no significant learning curve for the capacity factor, in that recent-lygtalled units have not improved in capacity factor over the older reactors.(Council on Economic Priorities, Power Plant Performance, N.Y.;Council on Economic Priorities, l976, p.I8.)See Enclosure~5 For more detailed discussion on the inclusion of all reactors in commercial operation of commercially viable size.

4 4 I)t'I', t.~~

l To W.J.Dircks~~V Page 4 4)As you well'know, the safety problems in nuclea'r reactor's have not been solved.t In Power Plant Performance, which is probably the best study on capacity factors in the nation, the Council on Economic Priorities projects that the average capacity factor for a l300 MWe pressurized water reactor (PWR)will perform at a 42.6%capacity over the first ten years of opera-tion.The study also concludes that the capacity factor of PWR's declines 3.2I%per l00 MWe incre'ase in size;thus the Palo Verde reactors would be projected to obtain a capacity factor of 43.563/o for the first ten years of operation.

Margen and Lindhe, two Swedish engineers, project that the capacity factor of the average plant in the U.S.in operation today will obtain an aver-age capacity factor of 42.7 over their lifetimes.(Peter Margen and Soren Lindhe,"The Capacity of Nuclear Power Plants", Bulletin of the Atomic Scientists, Oct.'75, p.40).They project that the capacity factor of the average plant will decline to 25%in a graphical line.Using the approximate slope of the record of capacity factors of nuclear plants in l973-74 (as the age increases, the capacity factors decrease)(Encl.<<6)-I have estimated the average foreseeable lifespan for the three reactors at the Palo Verde (Encl.<<7).Using your figure of$2.3 million for one unit s decommissioning, the cost of mothballing all three units of Palo Verde would be$8~.6 million including contingencies.

Furthermore, the Palo Verde reactors are to be the largest in the U.S.upon completion.

The annual upkeep., expense was not stated in the letter to Mr.Rhodes.I personally question the wisdom of using as a source the Atomic Industrial Forum.History shows that this forum has exaggerated the benefits of nuclear power con-sistently.

Your letter did not discuss the fact that experience in Europe (see Encl.<<>I)shows that dismantling may cost much more than projected by the Atomic Industrial Forum.With the exception of the error on construction cost of the Palo Verde ($4.35 billion should be replaced by the minimum cost of$3.78 billion), the cost projections for the report,"Palo Verde Economics-APS Projections Versus National Averages" have been found to be accurate and perhaps even a little low for nuclear power.I believe the efficiency projections (fuel efficiency and capacity factor)made in the report will withstand any kind of scrutiny.I would like to conclude with a quotation that will undoubtedly become well-known in future.In Mr.Saunders Miller's new book, he says"from an economic standpoint alone, to rely upon nuclear fission as the primary source of our stationary energy supplies will constitute economic lunacy on a scale unparalleled in recorded history, and may lead to the economic Waterloo of the United States." Sincerely yours, Russell J.Lowes Arizonans for Safe Energy 6I8 N.Central Ave.Phoenix AZ 85004 cc: The Hon.John J.Rhodes W'4 4 4 lI IW 4'~Ir I~)'~/4 4 I~II l l 1$'(~4 ,.I I'4 4 PALO VERDE ECONOMICS-APS PROJECTIONS VERSUS NATIONAL AVERAGES Submitted to the Arizona Corporation Commission

+ill.g~~~n7v r Russell J.Lowes~~or.rizonans for Safe Energy 6I8 N.Central Ave.Phoenix, Arizona 85003 I P a Ii H I'i II

~~Chapter I APS'CONOMIC PROJECTIONS FOR THE PALO VERDE PLANT The following report will deal with Arizona Public Service (APS)economic projections for the Palo Verde Nuclear Generating Stationi in contrast with national averages and trends for reactors already in operation.

Up to this time APS has never released to the public a thorough comparative report on the costs of nuclear vs.coal energy.For the Palo Verde, APS has made an estimation on the economics based on studies done elsewhere in the nation.The com-pany projects that electricity From the Palo Verde plant will be 38 per cent cheaper than coal-fired electricity.

They project nuclear at 40 mills*per kilowatt hour, and therefore coal at 64.52 mills per kilowatt hour.APS'rediction for nuclear generated electricity is based on false assumptions to such a degree that nuclear is made to look economically better than coal, when in actuality coal energy is more economical in Arizona.lt should be assumed that the cost for coaI<ired electricity as projected by APS's fairlyaccurate.

The company has been building coal-fired plants for years, and is presently involved in construction of such plants.ln order to conduct a comparative economic study on nuclear energy, there must be a breakdown of the costs, and a computation of the energy output.The utilities'ajor cost categories directly concerning electrical production from the Palo Verde plant are*I mi I I equals I/l0th of one cent.Arizona Corporation Commission, Arizona Public Service Rate HearincC Transcript, (Phoenix, Arizona: Hardy V/.Scott 8 Associates, March 1977), p.I222.

V V VV~~II II~vtI V I'V V'VN,it*V I V~'V VV'VI-~I V\I VI II~11 I,I Vl V t!'V 15~V V~$

~~capital investment, fuel costs, operation and maintenance, and decommissioning (see Table I).Then there are hidden costs such as waste storage;police protection of the wastes, fuel in transport, and the plant;and research and development-mf which the utility pays only a small part.The electrical output must then be figured.This output is expressed in terms of kilowatt or megawatt hours of electricity (kWhe or MWhe).Although APS owns only a portion of the Palo Verde plant, this report will deal with the economics of the entire plant.Interest charges will be neglected, as will taxes.The plant life will be assumed at 35 years.(The Final Environmental Statement by the Nuclear Regulatory Commission states the plant life at both 30 and 40 years.)Table I Cost Category~~APS PROJECTIONS FOR PALO VERDE G ENERATED ELECTRICITY (over a 35-year lifetime)Mills per kWhe'otal Cost (billions$)Capital investment Operation 8 maintenanceb Fue lc.Decommissioning 3.I7 4.0 5.89*(not figured)2.78 3.5 5.I6 SOURCES: Final Environmental Statement-Palo Verde Nuclear Generating Station, Units I, 2, and 3, NUREG-75/078 (Washington, D.C.: Government Printing Office, September 1975);Arizona Corporation Commission, Arizona Public Service Rate Hearing Transcript(Phoenix, Arizona: Hardy W.Scott 8 Associates, March l977, p.I222).Ted Dando, Nuclear information Representative for Arizona Nuclear Power Project;personal letter, March l8, l977.5.89 mills is the estimated average for the year l990.Estimations after l990 are unavailable.

~y~~II I 8 C IP M 7 I g II,

~~Costs for operation and maintenance (08 M)and the'hidden costs will not be dis-cussed in detail in this report.08M figures are unavailable on a nationwide basisg how-ever the"l9th Steam Station Cost Survey" showed that 08 M costs of nuclear power are slightly higher than coal.l Leonard M.Olmsfed, ed.,"l9th Steam Station Cost Survey", Electrical V/orld, l5 November l975, p.47.

I~P 4)~I+r Chapter II PROJECTIONS FOR CAPITAL INVESTMENT Capital investment for power plants is the amount of money required to build them.The capital costs for a nuclear plant are higher than those for a coal plant.l Arizona Public Service projects that Palo Verde will cost$2.8 billion;or$730 per kW (installed).

Construction of the nuclear plant probably could not have been finished at that cost today.The average construction cost per kW in l976 was$773 for plants completed tliat year.The Palo Verde should cost more than the average plant because cooling towers are being installed.

Cooling towers add an additional

$85 per kW as compared to the average cooling system.Nuclear plant capital investment has increased on an average of fifteen per-cent per year since l965, while in the same time the cost of building a coat plant has gone up less than ten percent per year.Even the Atomic Energy Commission (which usually underestimated the costs of nuclear energy, compared to the national averages)increased nuclear capital projections 500 percent from l968.4 Leonard Me Olmsted, ed.,"l9th Steam Station Cost Survey," Electrical World, 15 November l975, p.47.Edward Coven,"Economics of Nuclear Power Are No Longer Optimistic," New York Times, l8 July 1976, sec.4, p.6.James J.O'onnor,"Why industriais Must Favor Caai," Power, September l976, p.7.Marvin Cooke and Mike A.Males,"Analysis of Public Service Company's Pro-Iecfions for the Black Fox Nuclear Stations", presented to the Oklahoma Public Service Commission, Tulsa, Oklahoma, 22 August l976, p.2.

h p I~,hth>>'x~'~~>>~h If the average escalation rate continues throughout the building period of the nuclear station, assuming there are no schedule delays, the cost per kW would total$2365;the plant cost would total$9 billion--3.2 times APS'stimate of$2.8 billion.The fifteen percent escalation figure may decrease substantially in the future, but the costs are definitely expected to soar.Edward Cowan of the New York Times writes, The unwillingness of nuclear engineering companies to promise delivery of a$l billion to$I.5 billion plant six or seven years in advance for a fixed price is symptomatic of the runaway economics-an unanticipated surge of capital, labor, and uranium costs-and other shocks that have buffeted the nuclear power industry in the last few years.I The Bank of America predicts that kW costs for nuclear plants will range from$I620 to$I907 in l985.2"E c o nomics of Nuclear Power Are No Longer So Optimistic", New York Times, l8 July l976, sec.4, p.6.2John Berger,"Nuclear Power-No Solution to Energy Crisis", Sari Francisco, l976 f p.8.(Mimeographed)

.

~t I II Chapter I I I FUEL COSTS Fuel Performance ln obtaining an accurate picture of uranium costs for a nuclear plant, a net energy gain per unit of uranium must be'projected.

Then the cost of the fuel must be projected.

The energy gain for nuclear fuel is expressed as millions of kWhe per ton of milled uranium, known as yellowcake.

There has been no official governmental wide-scale sur-vey of this energy gain, but there have been estimates and small-scale surveys.APS is projecting a lifetime use of 48I9.62 tons (based on the assumed 35-year lifetime of the plant)of uranium dioxide (UO2)..There is a conversion ratio of I:7.3f of UO2 (which is the finished product)to yellowcake, with a margin of error.In other words, for every pound of UO2 used for fuel in a reactor, it takes 7.3 pounds of yellow-.cake (approximately).(NOTE: The reason that this step was used, instead of just using APS'igures on yellowcake, was because the author has been unable to obtain such fig-ures from the Public Service Company.)Multiplying 48I9.62 tons by 7.3 yields 35, I83.226 tons of ye llowcake.Dividing the total APS estimation of energy output by the amount of uranium to be used gives a projected 25 million kWhe per ton of yellowcake.

U.S.Nuclear Regulatory Commission, Final Environmental Statement-Palo Verde Nuclear Generating Station, Units I, 2 and 3, NUREG 757078 (Washington,D.C.:

Government Printing Office, September l975), sec.3, p.4.2Phone interview, Jim Harding, Special Advisor, California Resources Agency, Sacramento, Calif., 9 March 1977.

I' According to M.C.Day, professor of chemistry at Louisiana State University, l4 million kWhe per ton of yellowcake is about the average ratio.l In a report to the Bul-letin of the Atomic Scientists, civil engineer Ralph G.Kazmann and Joel Selbin, profes-sor of chemistry, apparently agree.They state that omitting such plants as the'Yankee Rowe and the Dresden 2 (which have poor performances), the average output is about l8 million kWhe per short ton of yellowcake.

Fuel Costs and Availability To accurately project costs for uranium, the supply must be determined.

The supply is nearing depletion on a worldwide as well as a nationwide basis;ie., the supply that is reasonably obtainable.

When the amount of uranium reaches a certain dilution, the energy required to process the ore outweighs the amount of energy obtainable.

Before this dilution occurs, the volume of earth to be mined becomes economically and environmentally unfeasible.

Tennessee Shale is such an example.It has been considered mineable by the U.S.Government.

Yet this shale has a very low energy gain.To obtain theaquivalent gross amount of electrical energy, it would be necessary Jo move about 2.3 times as large a vol-ume of uranium ore as coal.The energy per short ton mined would be 980 kWhe for ura-nium, and 2,250 kWhe from bituminous coat.Uranium is running short in the United States, and this is the type of reserves the suppliers will have to turn to.'M.C.Day,"Nuclear Energy;A Second Round of Questions", Bulletin of the Atomic Scientists, December l975, pp.53-54.Ralph G.Kazmann and Joel Selbin,"Letters", Scientific American, April l976,P.S.M.C.Day,"Nuclear Energy: A Second Round of Questions", Bulletin of fhe Atomic Scientists, December l975, p.58.

W I 1 i l I u F I But what about finding more uranium?There have been no major uranium deposits identified in this country in the last seventeen years, according to Robert Ninninger of the U.S.Geological Survey Uranium Branch.William C.Carley of the Wall Street Journal writes,"Geologists are especially concerned because they think most of the easy-to-find uranium deposits near the surface have already been discovered.

"2 M.A.Lieberman, associate professor of electrical engineering and computer sci-ences and a member of the Energy and Resources Group at the University of California at Berkeley, estimates that the'total amount of uranium that can be assumed to exist in the United States is I, I34,000 short tons.He concludes: "It will be shown that, if the expan-sion of nuclear power proceeds as planned, a serious shortfall in uranium will develop during the late l980's." Most of the uranium in the U.S.comes from two geological formations known as the Colorado Plateau and the Wyoming Basins.We will need to discover new uranium supplies equal to nine Colorado Plateaus or twenty Wyoming Basins.Hans Adler, Geolo-gist for the Energy Research and Development Administration Nuclear Fuel Cycle and Production Division, states: "The ma jor question confronting exploration geologists, is where in the U.S.will facsimiles of these two regions be found once, much less 9 or 20 times." Ralph E.Lapp,"We May Find Ourselves Short of Uranium, Too", Fortune, Oc-tober l975, p.I5l.William C.Carley,"Uranium Drain.Fuel Shortage Forecast for U.S.Nuclear Plants Within Decade or Two", Wall Street Journal, 7 June l976, sec.I, pp.I, IO.M.A.Lieberman,"United States Uranium Resources-an Analysis of Historical Data", Science, 30 April l976r p.435.David Dinsmore Comeyr"The Uneconomics of Nuclear Energy", S~he tic, July l976, p.2I.

~l A Utility companies are unsure of their future sources.The supp!iers are not consent-ing to long-term contracts, unless there is allowance for them to increase costs in the event their costs go'p.Such is the case in APS'ontract.

Commonwealth Edison of Chicago, the nation s largest nuclear utility, admitted it had no idea where its uranium would come from after l980;its fuel manager could only say,"We must believe the resources will be there to keep those monsters running." There have been large companies defaulting on contract prices already.Westing-house, the nation's leading uranium supplier, found itself unable to provide 50 million pounds of natural uranium to utilities at the contracted prices and defaulted, prompting lawsuits.APS has contracted with Westinghouse for its uranium, even though the supplier will find it hard to meet other contracts throu'gh the l990's.Other companies have also announced deficits, such as General Electric, the second largest supplier in the U.S.To top it all off, there is now an OPEC"like uranium cartel that is expected to purposely jack prices up.4 What about the Breeder-or Recgcling'?

l t ha s been rumored that the br'ceder reactor will relieve resources requirements for uranium.That would be partially true if the breeder program is okayed.Many conserva-Ariz.Corporation Commission, APS Management Study(Phoenix, Arizona: Peat, Marwick 8 Mitchell;inc., l976), sec.4.6, p.lTT4.2 William J.Lanouette,"Nuclear Fuel: Will lt Run Outl", National Observer, 24 April l976, p.l.'Westinghouse:5 the Waiting Period", Forbes, I December I975, pp.24"35.4"It Worked for.,the Arabs", Forbes, I5 January 1975, pp.I9-21.

~1 E'4 1 n 4 I*J I f!4,~

IO tive political leaders who support the current nuclear program are opposed to the breeder.If the program survives, the cost is expected to be tremendous.

A General Accounting Office report, revealed by columnists Jack Anderson and Les Whitten, states that if the program comes on linerif auld cost$I53 billion to build the same energy capacity that could be constructed for$128 billion with conventional reactors and$95 billion with coal-fired power plants.I If the nation turns to fhe breeder, it would have no significant effect on uranium requirements in the year 2000, according to Dr.Ralph Lapp, a breeder proponent.

The program would possibly depress uranium prices later on.The situation on recycling is similar;it.just may never be gotten around to.In this case too, there are dangers involved.But, if recycling becomes a reality, the most that the uranium supply could be boosted would be 50 percent, but probably less than 25 percent, according to M.C.Day.Because of the shaky grounds on which the breeder and the recycling program stand, because of the complications in implementing either, and due to the lengthy lead time in constructing either type of plant, it is evident that even both together will not relieve our uranium shortage.Jack Anderson and Les Whitten,"Secret GAO Report gives Nuclear Energy Dim Look", Scottsdale (Az.)Daily Progress, 3 November t976, p.4.Ralph E.Lapp,"We May Find Ourselves Short of Uranium, Too", p.I99.M.C.Day,"Nuclear Energy: a Second Round of Questions", pp.53-54.

J I J Chapter IV CAPACITY FACTORS In figuring the costs of power from any plant, a comparison must be made between how much money is put into producing the electricity and how much electricity is'delivered.

I What is actually produced from a plant is referred to as electrical output, and Is expressed as a percentage of the amount of electricity that could have been produced, had the plant been in perfect running order at all times.This percentage is called the"capacity factor" of the plant.To be more specific, the capacity factor of a power plant refers to the number of kWhe a plant produces in a given amount of time, divided by the number of kWhe that the plant could have produced, if the plant had been operating l00 percent of the time at full performance.

If two plants require the same cost for'building, operating, maintaining, and fueling, but one has an average capacity factor of 80 percent over the life of the plant, while the other has an average of 40 percent, the first plant generates electricity for one-half the cost required by the second plant.Palo Verde will have three separate reactors, each producing I270 MWe at full performance.

APS has predicted an average capacity factor of 75 percent over the life of the three reactors.This projection is totally.unrealistic.

There has never been a reactor over l000 MV/e in size to operate at this high percentage for a full year.The larger the reactor, the lower the average capacity Factor will be.The Palo Verde reactors are to be the largest reactors in the United States.

ln the U.S., the average nuclear plant size is between 700 and 800 MWe.Two Swedish engineers calculate an average capacity factor of 42.7 percent for fhe average size reactor.I Evidence of the lower capacity for larger plants exists in the record of actual per-r formance of nuclear reactors.Jim Harding, Special Advisor to the Energy Resources Con-servation and Development Commission of California, has calculated the total lifetime commercial reactor capacity factor to be 53.7 percent.The average capacity factor for plants over I000 MWe is about 44.5 percent.It is not likely that capacity factors will improve for plants over l000 MWe.In fact, the cumulative-to-date capacity factor average for plants over l000 MWe went down from 46 percent in l974 to 44.5 percent in l975.There has been no significant improve-ment-or learning curve--for nuclear performance since l973, the first year that the U.S.Government started releasing reactor performance records.Furthermore, a retrogression

',*";of capacity factors is expected'.

All plants'over l000 MWe in the U.S.are less than eight years old, and after the eighth year of operation, capacity factors decline throughout the rest of the reactors'ifetimes.

There is not enough statistical information available to give more than a general estimate on the capacity factor that nuclear reactors will decline to, by the time they are shut down.But they are expected to go down to about 25 percent for the average plant.6 David Dinsmore Comey,"Points Vs.Trends", Bulletin of the Atomic Scientists, Oct.75, p.45 Jim Harding, personal letter, Special Advisor, Calif.Energy Resources Conservation.8 De-velopment Comm., 8 February l977.David Dinsmore Camey,"No Improvement, Capacity Factors Stay Constant in l975", Not Man Apart, March 1976, p.II.4 Ibid.5Charles Komanoff, Power Plant Performance (New York: Council on Economic Priorities,: I976), p.4.Peter Margen and Soren Lindhe"The Capacity of Nuclear Power Plants", Bulletin of the Atomic Scientists, October I978, p.40.

~I I j j I I3 The Council on Economic Priorities, a consulting firm based in New York, did a detailed study on nuclear and coal capacity factors and projected performances for a range of different sizes of reactors for the first ten years of performance (see table 2).Table 2 LEVELIZED AVERAGE PWR CAPACITY FACTORS Ages I-IO Unit Size 500 MWe 600 700 800 900 I000 I I 50 t300..K9..66..62..59..56..52..47.2.8.7.6.6 Projected Capacity Factor S OUR C E;Charles Komanoff, Power Plant Performance (New York: Council on Economic Priorities, l976), p.32.*PWR is the abbreviation for pressurized water reactor, which is the type being built at Palo Verde.

Char t'I'V DECOMMISS IONI NG The costs of decommissioning have almost been ignored by APS, and certainly have not been figured into the total co't.There are several figures for decommissioning costs that are circulating.

The most common figures are$1 m)llion, plus$I00,000 per year indefinitely.

Indefinitely, i'ndeed!"Because of the very long half-life of nickel-59, exposure from gamma rays and X-rays from this source in a commercial reactor would not decline to the permissible level of 0.2 millirems per hour in a 40+our week for l9.28-80,000 year half-lives, or I.56 million years," according to the New York Public Interest Re-search Group at the State University of New York at Buffalo.Decommissioning has proven to be much more expensive than most utilities estimate.The Elk River reactor, which was 22 MWe as compared to the three!270 MWe reactors of the Palo Verde, cost$6 million to construct and$6.9 million to dismantle.

This ratio should not be casually scaled up to present%ay prices, but dismantling is expected to cost much more for commercial sized reactors.The Public Interest Research Group stated that dismantling will certainly amount to tens of millions of dollars per reactor.S t e v e n Harwood et al,"The Cost of Turning It Off", Environment (December l976), p.I8 Jersey Central Power and Light, a New J rsey utility company, is seeking per-mission from the state's Board of Public IJtility Commissioners to boost its rates so it can start building a$100 million fund for the purpose of decommissioning a nuclear power plant.The plan is to raise$I.35 million a year, to be set aside in the Form of tax-free government securities.

This would raise$I00 million by 2033.The funds that this utility is seeking to obtain may be well under the amount re-quired for decommissioning.

In reference to the costs at the Elk Piver reactor, Chemical and Engineering News reports;Similar experience in Europe indicates that the cost of this procedure runs about 45%of the value of the initial investment.

In any event, there are no unique technical problems associated with taking a facility out of service.>Because a large reactor has never been decommissioned, it is hard to tell how much the price should be scaled down."In Place Entombment", Stevens Point (Wis.)Daily Journal, l4 January I977 f Sec.I, p.4."Experts Mull Over Radioactive Waste Disposal",Chemical and Engineering Views, 2 August 1976, p.23.

Chapter Vl H1D DE N COSTS There are many hidden costs of the Palo Verde.The government will eventually carry the burden of storing the high level wastes, and will probably share costs of de-commissioning.

There are other government costs that have already indirectly gone into the Palo Verde plant.Committee for Nuclear Responsibility has estimated"For each nuclear plant licensed to operate so far (about sixty plants), taxes provided almost$100 million in gov-ernment research and development." For 1977 alone, the Nuclear Regulatory Commission will receive a quarter of a billion dollars in tax money.2 None of this money is for the military nuclear program.The Energy Research and Development Administration will spend around$5.8 billion for their total nuclear program, some of which is for commercial purposes.The water requirement for a nuclear plant is much higher than the requirement for a coal plant.The Palo Verde will require 75,000 acre-Feet per year.A coal plant of imilar size requires 45,000 acre-feet per year.Since Arizona is not getting all of the power from the nuclear plant, much of Arizona's water will be used for out-of-state power.Arizona will lose a tremendous amount of water, without a fight."Nuc lear Power-Bad for the Economy": Committee for Nuclear Responsi-bility, Inc., Yachats, Ore., Nov.15, 1976.2 Executive Office of the President, The Sudget of the United States Government, Fiscal Year 1977: Appendix (Washington, D.C;:&overnmentlrinting Z)7fice, lag, 16 V

Chapter Vll PROJECTIONS BASED ON EXPERIE NCE APS'laims are not supported by the facts documented in this study.Taking into account the data in Table 2, and the expected decline in capacity factors, it appears that the average capacity factor for reactors the size of Palo Yerde's will average about 35 percent over the lifetimes of the reactors.This is 2.l4 times lower than the estimate given by APS~Concerning fuel, it would appear that because of the lowered capacity factor, the amount of fuel purchased should be less than one-half the amount that APS originally projected, thus cutting costs.This, however, is not the case.The apparent saving is nearly cancelled by the fact that the fuel efficiency ratio claimed by APS is 75 percent higher than the actual national average.It seems obvious that yellowcake will go into the hundreds of dollars per pound, and will cost at least$200 per pound by the l990's when our mineable domestic reserves ,I near depletion.

At this minimum price, the cost for fuel would be 28.57 mills per kilowatt-hour of electricity, instead of the APS claimed figure of 5.89 per kilowatt-hour.

Noting that the costs of construction have gone up drastically and are apparently continuing to climb, this would indicate that the costs of the nuclear plant will finally run between$4.35 billion and$9 billion-assuming that the inflation rate does not in-crease.Based on the projected costs of decommissioning the plant in New Jersey, the cost I M.C.Day,"Nuclear Energy: A Second Round of Questions", Bulletin of the Atomic Scientists (December l975)p.54.l7

~~1 l I8 of, the Palo Verde plant's decommissioning (because Palo Verde is to be the largest plant in the nation)will probably run to over$200 million--the equivalent of 0.49 mills per kilowatt-hour of electricity.

These total costs have been grossly understated by the public service company (see Table 3), as the capacity factor has been overstated.

Table 3 POSTULATIONS VERSUS AVERAGES AND TRENDS (Over a 35-year L i f e t i m e)Cost Category Mills Per kWhe A P S Probable Total Cost (Billions$)APS Probab le Ca p i ta I investment*

3.I7 l0.63-22.02 Fuel 5.89 28.57-7I.43 Decommissioning (not figured)0.49+2.78 5.I6 4.35-9.0I ll.69-29.22 0.2+"disregarding extra cost for cooling towers Consolidating APS'xaggerated projections on investment input and electrical output boosts'he costs per kilowatt hour from both'ends.

Subtacting APS'stimates on capital invest-ment and fuel cpst from the minimum probable estimate yields an increase in cost of 8.3 billion dollars.Dividing APS projected electrical output into 8.3 billion gives 9.467 mills per kWhe, in additional costs.This gives 49.467 mills per kWhe at the estimated 75 percent capacity factor.The above figures concern only capital investment, fuel, and decommissioning costs-they do not include operation and maintenance, or interest.Using the 35 percent capacity factor boosts the cost of electricity 2.l4 times to I06.0 mills/kWhe.Nuclear energy from the Palo Verde plant will cost at least I65.0/c more than APS has been projecting.

With coal at 64.52 millsAWhe, nuclear energy will be at least 64.29 percent more expensive than coal-fired electricity.

1~

REALISTIC URANIUM ENERGY YIELDS AND COSTS by Ron Carstens, Robert Lamson Nuclear power plants haye always b en known to have high"front end" capital l costs, and since l970 the escalation in capitql costs has posed an increasingly serious economic problem for nuclear power.However, nuclear power advocates in government and industty have attempted to counteract these high capital costs by empha:izing the"practically negligible" uranium fuel costs of nuclear power.The uranium fuel cost component oF electricity produced from nuclear power'has received scant attention outside of cost estima'res for specific situations and'individual"power plants.In almost all cases, these"official" costs for the uranium fu" I component have been around 3 mi Is p r I wh J I I with recent figures put at 5 to 6 mils.'tility industry official have b"en led to believe that this v,'as)he primary advantage of nuclear over coal pov er and offset nuclear's highc-capital costs., However, there seems to be a lack of'ublished op rating na!a on whicn to base such fuel cost calculations, particularly electricity yield from uranium.On fhe other hand, l,4,!2'here is a great deal of information published Gn exp'ecled or projec(ed yields.'I fhis investigation was undertaken to establish v~hat uranium fu" I cost co,nponent utilities could exp cI for a nuclear power plant in the near future, b"sed upon actual operating data.All ulililios operating nuclear power plants for more than five years were invited to submit their uranium fuel loading record, including dates lt fuel charged and i)s enrichment.)Ye obtained data from startup for and tne ten-year recorc.of two other plant~a from ERDA.The total I I IO I of!oading, amount of five utl I)ties (5g 6g 7g Bg 9)e Ieclrica I generation for these plo>>ls from slailup date to tive latest fuel loading dole was derived from data in I

~s~~'the Federal Government publication"Monthly Operating Plants Status Report" known as the"Gray Hook." Utilizing the fuel loading record and its enrichment, the uranium oxide"yellowcake" required to fuel each plant was calculated from a material balance on the enrichment stop, using urapium ores naturally occurring 0.711 wt%U23J" in the feed and a tails assay of 0.3 wt%.'The fuel'charges used were the sum of the individual 1,4,11 loadings charged to the plant from startup until the Fuel cycle just prior to the last one submitted by the utility.In this way, the electricity generated from the last fuel load charged to fhe plant would be included to give maximum credit for generation fro'm this last load.These overall electricity yields, as summarized in Table I, were simply calculated by dividing the total electricity generated over the periods described by the net fuel charged converted to uranium yellowcake (100%U308 bali:).The individual plant's yields were, then averaged b>weighting ll>e yield v;ith lhe opera/ing life to give more weight to the~.plants operating the longest in order to de-emphasize the yield-damp ning effecl of the inilial core loading.The average yield obtained was 12.36 Mkwh per short ton yellowcake,~/~~~~which is aslonishingly low by previous statements, being less than 40%of the lowest previous 1,12 official government and industry statements.

In o'.Ficial Federal Government publications

's well as private'nd industry manuscripts, the energy yield From uranium is represented 413.3 as"unchullcnrucatstc ansi immutabie" at arcunst 32r600 Mcrtaratt (thermal)days,per metric'on uranium metal in the fuel.This calculates out to something over 32 MMkwh per short ton)rellowcal;e over a ten-year operating period williout rep'roccssing, depending upon reactor typ., etc.Additionally, in l970 ilia AEC stated that reactors yielded 34 MMkwh I per short ton yellowcake, without rcprocessing.~

Just last year,an ERDA official teslified s that light wolor nuclear reaclors in)lie United States routinely contributed 32 MMkwh per 7 t short ton ye llowcake to the Vni ted States energy needs (without reprocessing).

Some 15 have claimed over 60 MMkwh per short ton yellowcake as the energy yield.On the basis of the data revealed here, it appears that present generation light water reactors may use at least two-and-one-ha'ff times as much uranium fuel as has been heretofore assumed by Federal planners.As can be noted in Table 1, the electrical yields vary a great deal, with the highest boing over twice the lowest yielding plant.Attempts were made to correlate these data with capacity factor, plant size, plant type,%enrichment, and between short and long-operating plants, all with no result, Capacity Factor would be a logical correlating factor since the plants are charged with a designated amounl of fuel and the rods are regularly changed even though the electrical output may be below design for various reasons, However, there is no correlation with any factors these authors could identify, We also considered the relatively brief operating time for the plants considered here (average life 7 years)with respect to the dampening effect of the initial core loading.However, using the e Government's energy yield figures for the initial core and replacem nt loadingsI we calculated the yield difference between a plant op rating seven years (pverage for this study)and The difference revea led here, those operating 15 years, or one-half the plant's expected operating life.in yield obtained was only 3%and thus would not account for the differences Meager published data on BWP.fuel rods performance would seem to be in agreement wiih this data.ActuallyI these electrical yields as derived here are high by 6%t'o 10%because no account was taken oF transinission lo"'ses which are real due to the large and remote nature o', these plants.Therefore, for a utility that is concerned wi h buying uranium yellowcako and delivering"billable" electricity to its customers,!he 1\f actual electrical yield based upon these operating data is well below l2 MMkwh per short ton yellowcako.

This analysis, of necessity, did not include the effe'ct of recycling I uranium or plutonium as those are not commercially practical at present.There is considerable doubt if there ever A(ill bo commercial reprocessing and recycling.

Even given the unlikely event of recycling becoming a commercial reality, it would at best o ol2 increase the yield 20/o to 25/o which may never be realized due,to the doubtful economics~IS of reprocessing and recycling.

In any event, it is questionable that reprocessing and recyc-ling can be operational fast enough to help the United States from exhausting domestic uranium reserves in the near future.Utilizing the electrical yield derived herein, the official ERDA figures on proven uranium reserves, and tho projected installation of nuclear plants, the United States l9 l2 could theoretically run out of uranium fuel for its scheduled plants well before l990 assuming no imports.Thus it is peculiar indeed that the nation is being asked to go nuclear in order to be self.ufficient in energy.In a day when the United States has problems with petro-dollars and an Organization of Petroleum Exporting Countries, we can expect to have problems tomorrow wi/h uranodollars and an Organization of Uranium Exporting Cou:itries.

fo The riext portion of this study considered the effect of this lowered eloctrical yield upon tho fuel cost component of nuclear generated electricity.

There are many costs associated with uranium fuel which are not considered in tradilional cost calculations for~individual utilil los'lants, These ossociated costs--spent fuel reprocessing to ultimate 0 waste management'nd subsidies on reactor development and'fuel enrichment--have caused some nuclear proponents to soriousjy re-evaluate cost ostimalos.

Idowever, for purposes 20 or this discussion those,posts are not, considered hero, and we restrict our analysis to the

fuel cost component strictly from the standpoint of an individual utility operating a nuclear fueld generating station which will be on stream in I985.The components of this fuel cost reflect only the operations of fuel gathering and preparation in order to prepare the fuel rods for light wafer nuclear plants.These include the cost of uranium yellowcake itself, transportation and conversion to UF6 for fuel enrichment, the enriching process (to allow the natural 0.7I1%U235 to be upgraded to about 3.2%, reconversion to U308 and fabrication into fyel elements to be placed in a reactor.The cost of yellowcake has undergone enormous increases in the past few years, fram a market price of just 7$/Ib in l973 to the present contract price of 40$/lb.Current long-term uranium contracts are written tied to market price at delivery or 7%per year escalation, whichever is greater.Therefore, the future price of uranium cannot be less than present prices (about 40$/Ib)plus 7%per year escalation.

This yields a l985 price of about 80$/Ib which is as low as a utility can expect to pay.Others have anticipated even higher prices,'ome wc II over l005/Ib for the m<<l980's.The~'2 23'I uranium supply situation is so serious that some utilities have been forced to invesl in 24,25 uranium mines as a defense against further escalation.

'urther aggravating the situation will be the lower-than-expected electricity yields observed here.At" the enriching plant step,.we can also see the effect of inflation and rising electricity prices on the troriiendous amounts of power it takes to enrich the uranium.Even considering ERDA's heavily-subsidized operation and using tax-free TVA power, the" Gave rnment charge for enriching has climbed I'rom$28.70/I:gSWU in I97l to the recently announced$67.25/kgSV/U which is almbst a 20%annual increase.Further cost increases will be forllicoming due to the increasing cost of electricity, lack of added~5 I

~~Government subsidy or private enriching plants, and the normal forces of inflation.

5'ith an escalation rate of 7%per year for inflation, costs double every ten years.Therefore, we estimated a 1985 enriching cost of$135/kgSNU, which is probably on the low side given past and future exacted electrical cost increases, the real costs of~government subsidy, and/or a shortage of enr'iching capacity.The final step oF recon-version of UF6 back to the oxide and fuel rod fabrication has been.est imat d at$100/kgU,~~~~~27 but'realistic estimate puts the figure at$125/kgV, which yields an escalated figure~For 1985 of$250/kg U..Combining these cost elements with the average operating electrical yield derived t in Table I gives the uranium fuel costs in Table ll.These 1985 basis costs total 19.8 mils per kwh just for the uranium fuel component alone into nuclear power plants.This is some 400/o to 600~o higher than previous estimates by government

nd industry, 0 O~~~12 34 I These costs do not include nuclear fuel financing charges which will add at least 20%and as much as 50%to the Table ll figure oF 1.9.8 mils per kwh.Financing costs vary greatly and depend upon actual vs.exp cted fuel rod life, how much fuel reserve.a utility likes I.o have and their financing costs.Any economic advantages which may come t'o'pass from reprocessing or"breakthroughs".

In enrichment technology will be completely overshadowed by these financing charges and by transmission line losses/which were al;o not included.Therefore there is reasonable assurance that a utility in 1985 will have pay al least 20 mils per kwh for nuclear fuel costs alone.lt would appear, based upon this analysis, that the nuclear Fuel costs pcss-throughs of)hc,1980's will malo lliose of the'70's seem low indeed.Fur)hermore the I~

highly touted fuel cost advantage of nuclear power will very soon turn out to be a 28 disadvantage.

Even if we escalate the present average U.S.price of coal (516.90/ton) to 1985 (at?loper year, yielding 533.80/ton), and use an average energy yield of 3000 kwh/ton for coal, we can see that the 1985 fuel cost of coal power is only 11.3 mils per I;wh,'or 57'o of the uranium fuel cost component in Table Il.t The present lower charges for nuclear plant's'-fuel is evidently a short-lived utopia that will be shattered as we enter the 1980's.

t l I k~~Q 1~~~~~~~~~I~,~,-~)-~~~J~~~~~~~~~~~~~~-o~~~~~~~I II~~~~~~i~~~~~~~~~~~0~~~~~~~~~~~~~~0~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~t~~~~~~~~~~~~~

,I~r Q 4 l I 1

~~~~~4~+TABLE I I C OST

SUMMARY

URANIUM FUEL COST COMPONENT OF NUCLEAR ELECTRICITY

• 1985 BASIS Cost/Unit Units/short ton ye I lowcake Cost per short ton ye I lowcake Mi Is/ViWH 100%U308 ye 1lowcake Transportation and conversion to UF6 Enrichment 80$/Ib 7,30$/Ib U 135$/SWU 2,000 239 413.7$160,000 w 1,750 55,850 12.95 0.14 4,52 UF6 reconversion to U308, fabrication into fue I rods 250$/kg U 108.7 27,180 TOTAL.~~~~~~2.20 19.81 mi Is/kwh a-Present NUEXCO.contract price escalated7%per current contracts to 1985.b-Approximate es'timate includes$2.86/Ib U for conversion and 27</Ib yellowcake equivalent, for transportation to and from conversion and enrichment and final delivery, both escalated at 7%per year./fo~-At average 3.2%enrichment, requiring 3.80 kg SWU for U fuel enrichment, and 9.2 short tons I j U308 per metric ton U.I II d-At Table I average yields of 12.36 MMkwh per short ton U308 yellowcake.

1 e"'urrent market prices escalc ted lo 1985 at 7%per year.

I~e I f 8l BLIOGRA PHY I.ERDA-52 2.Final Environmental Statement NUREG-7S/053, May 1975, Skagit Nuclear Power Project, Table 10.2 3.Talk given by Gordon Corey, Vice-chairman Commonwealth Edison, Nov.17, 1975I at M.l.T., reported by Atomic Industrial Forum 12/8/75.Appendix B, Nuclear Fuels Policy Paper-1976, The Atlaritic Council of the U.S.I Library of Congress Card No.76-151-28.

5.Letter 12/IO/75 from Jim Karalus, Northern States Power Company.6.7.8.Letter 10/22/75 from George Fox, Northeast Utilities Service Company.Latter II/11/75 from Robert Mecredy,'Rochester Gas and Electric Corporation.

Letter II/3/75 from John Madgett, Dairyland Power Cooperative.

9.Letter IO/28/75 from Larry Smith, Carolina Power 8 Light Company.10.Letter 8/19/75 from Frank Baranowski, Director, Nuclear Fuel Cycle, ERDA, II.ORO-684, AEC Gaeous Diffusion Plant Olmrations.

12.YVASH 1139, Table 8.13.Paper from lecture in Japan by Manson Benedict, M.I.T.Nuclear Engineeringl April 1975 (Fuel Cycles for Nuclear reactors-Uranium Enrichment and Repr'ocossing, p.1.5)14.Letter from George Kavanaugh j General Manager for Reactors;AEC, to Senator Frank Moss 10/I/70.15.Testimony of John Patersonr Chief Supply Evaluation Branch, ERDA, at NRC hearings, May 1975, on River Bend Nuclear Units I and 3, St.Francisville, LA, 16.V/all Street Journal, Feb.17, 1976, page I.17.Testimony of P.Gans, Bonneville Power Administration in NRC hearings, Skagit Nuclear Project (June 1976).

~~v A~~Sa P l8.M.Resnikoff, Expensive Enrichm'ent, Environment, Vol.I7, No.5, p.28-35, l9.ERDA Weekly Announcements, April I6, 1976, 640,000 Tons Proven Reserves.20.Weekly Energy Rep'ort, March I, l976.2l.Nuclear Exchange Corporation Quotations, Early I976.4~I 22.Bellingham llerald, Dec.28, 1975, page'I>for Puget Sound Power&Light, Skagit Nuclear Project (83$/Ib).23.David Snow of Mitchell-l.lutchins, N,Y.C., October l975, Institutional Investor.24.Wall Street Journal, Jan.I9~I976.25, Syndicated Report of Commonwealth Edison uranium mi'ne purchase, by Bruce Ingersolg Chicago Sun-Times.

26.ERDA Weekly Announcements, PAar.5, l976.27.Privale communication with Dr.R, Bardes, Exxon Nuclear, 4/23/76.28.Federal Pov,er Commission News, Mar.4, l976.29.Nail, Stobbs,"The Nuclear Assurance Corporation Performance Program/" Symposium Proceeding, Vienna, Oct.8-I2, l973, International Atomic Energy Agency STI/PUB/35.

I I'1 gf~g-l a'I5,'N:o~4~p FLAK, l lgeS~rceS~+a'Aguji c~~+" Orang,.=8pr'/'6'~~.

P~~U.S.URANlUlVi RESOURCES AT$10 TO$200+'?PER LB.U30g 4000 5200i2 200 5100<5 I 00'5200 6 C7$X 550 5 I 0 0 CONYENllONAE SHAlE SHAEE GRANIIE SHAEE GRANITE SEAWAEER 60-$0 ppo 2S-60 ppm l0 20 ppn 10.2S ppn 4-10 ppn 0.00)ppaa FJgws C ea I 0 l I l 10 THE ECONOMICS OF NUCLEAR AND COAL POWER~~(~V URANIUM DEMAND.AND SUPPLy 11: 0 tB 0 0 tax O C A 0..IVith limited resources, a more cogent method of comparing supply,and demand might be, to calculate the number of reactors that can be fueled for their entire life cycles.For a 1,000 I~AVe reactor with a 30-year life, 5,069 metric tons (MT)are needed.This estimate is based on an initial charge of 545 MT and an annual reload of 156 MT (at a 65 percent capacity factor).Dividing the tons of uranium available by 5;069 will yield the number of 1,000 Mme reactors that can be fueled for their lifetimes." lt will-bc noted from Table 1.7 that, at most, 117 plants of 1,000-iiINe capacity can be supported for their, lifetimes by known domestic reserves.How-ever, as of September 1975, the Nuclear Regulatory Commission had already~At an eAicicricy nte for converting heat to electricity of 32.6 percent (known as 32.6 MtVDth/kg U), I kilogram (kg)of enriched uranium produces 258 200 kil h (w)or.2582 X 10 kwh.(MtVDih=thermal megawatt days.)Thusareactoroperating at 65 percent capacity would require 22,052 kg of ennchcd uranium per y:ar.(8760 hour0.101 days <br />2.433 hours <br />0.0145 weeks <br />0.00333 months <br />s/yr.)

(1,000 Mme)(65%Cap.Fac.)=(8.76 X 103)(106 KW)(.65)=5.694 X 109 KWH/yr.(5.694 x 109)/(.2582 x 106)=22,052 kg.At 030 tails assay 7.08 kg.of natural uranium (U308)are required to produce I kg.enriched un:dum.1,000 kg=I iiIT TABLE 1.7 Reactors Supportable by Uranium Resources 65 Percent Capacity with no Plutonium Recycle Number of 1,000 MWe Plants Supportable Resource Base Without MT U308 Reprocessing With Reprocessing'nown reserves Plus potential:probable Total Plus potential:

possible Total Plus potential:

specu-lative Total 543,000 1,034,000 1>577,000 1,215,000 2,792,000 372,000 3,164,000 107 311 550 624 117 341 604 684'If the reprocessing capability projected by ERDA does, in fact, turn out io provide 10 percent of the uranium needed,the life cycle demand is reduced to 4,620 SIT.According to D.E.Saiie (CONF-750209, pp.67-'72), ieprocessing is expected to provide from 1.4 to 23 percent of total feed requirements between 1976 and 2000.In reality, reprocessing capa-bility is far short of any projected targets, bui an average.of 10 percent is being utilized foi discussion purposes.Source: Compiled by the authors.t'7 0 0 0 0~I 4 tB~~C9 R cd CQ v T I I/~02,052 x 7.08)/(I,000)

=156'AT/ur.

of natural uranium Assuming,a30-year life for a plant, we have the following equation: (initial load)+(years X reloads)=life cycle requirements for one plant.545 MT+(29 x 156)=5,069 MT It should bc noted that an efficiency rate of 32.6 percent is an assumed conversion rate utiTized by ERDA in its computations, and our methodology has paralleled theirs.tvhether or not this high an cfficicncy rate is justifiable is open to question.In the Bulletin of the Atomic Scientists of December 1975, M.C.Day (" Nuclear Energy: A Second Round of Questions")cites testimony of F.B.Baranowski of'RDA as testifyin to S.F k E..oss on.idy l.1974, that the conversion ratio was 14 million kwh pei short ton o('ranium oude, compared with the more commonly used value of 32 million kwh per short ton.Day does iiot draw any defmitliv conclu>lons, but if the efliciency rate is indeed less than 32.6 percent, the ability of uranium resources io sustain iiucicar plants will be con-si embly less than indicated in this chapter which is based on ERDA assumptions.

granted operating licenses to 37,000 MiVe of nuclear capacity, construction per-mits for 65,000 MWe, and limited work authorization (ground breaking permits)for IS,000 MNe-a total of 120 Gee of committed nuclear capacity.Unless more"knout reserves" are found.some of these reactors may have to shut down before their economic lives are completed.

Prudence would dictate that no more new reactors be started until"potential resources" become additional"known reserves." However, ERDA projections show that by 19SS, between 43 and 128 plants will have been built, which, based on known reserves, will not be.able to function for their full economic lives.Table I.S illustrates this.In fact, all projections forecast plants coming on stream after the known reserves would be fully committed (were prodticers to sign 30-year contracts).

Under the moderate!hipo case (considered the-reference" case by ERDA)demand created by piants begun through 1990 would exceed all resources by 32 percent.It should be understood that the"potential resources" at.this time-"probable.""possible," and"speculative"-are only"maybe's."

~a tors at other plants.However, the suggestion that outages at these two units were not size-related but nevertheless cause the appearance of a size impact on capacity factor, is credible and worth considering.

The trend equations through 1975 are indeed sensitive to inclusion of data for the two units.The PMR size-and-age equation through 1975 has a size effect of 3.4Ãper 100 Hw, with 99.9X statistical.

confidence (one-tailed) for the size term.Dropping Palisades, the size effect falls to 1.8%per 100 Hw, with only 97%confidence (but with a goodness of fit of 27K, vs.21K for.the all-PMR equation).

For BMRs, the size effect through 1975 of 3.3X/IOO Hw falls to 1 6~/100 Hw without Browns Ferry I)l, with the confidence level-falling from 98K to 87K when the unit is excluded.Clearly the size ef-fect on capacity factor is.extremely sensitive to inclusion of Palisades and Browns Ferry 01 (althouth Coannnwealth Edison is incorrect in stat-ing that"there is no significant correlation of unit size"-to PWR capac-ity factors without Palisades).

Nevertheless, we believe it was proper to include these units'ata in the Power Plant Performance trend equations, for the following reasons.(i Exc usion o some ata opens the door to subjective decisions in defining the appropriate data base.If the worst performing PWR is removed from the sample, should Robinson 82, a reliably performing 707-Hw PMR, be similarly excised in a kind of trade?(Robinson it2 has the greatest cumulative positive divergence from the capacity factors pre-dicted by the trend equation for a PMR of given size and age.)The ar-bitrariness in such decisions is eliminated by including all the data.(if)It is likely that the duration of outages affecting Palisades and Browns FerryÃcapacity factors was size-related, even if initial outage causes were not.Large plants such as Browns Ferry fl have more complex instrumentation than smaller units, so it is likely that the Browns Ferry fire caused cable damage which was aire extensive and re-quired more repair work than a comparable hypothetical fire at a.smaller unit.Similarly, the s~or problem at Palisades-chronic steam gener-ator tube leaks-is probably somewhat size-related, since the number of-steam generator tubes is proportional to unit rating.It is thus not warranted to drop all of the size impact of the Browns Ferry tl and Palisades data from the trend equations.(ifi)A look ahead at 1976 half-year data during preparation of the CEP study demonstrated to us that not only was 1976 full-year data likely to increase the significance of the PMR and BMR capacity factor/size correlations, but also that the magnitude of the size impacts was becoming less sensitive to inclusion of'Palisades and Browns Ferry data.Sure enough, through 1976, the average size effect on PMR capacity fac-tors is 2.5X per 100 Yw without Palisades (with 99.9X confidence), and 3.3X/100 Hw with Palisades.

Through 1975, the size effect on PMR capac-ity factors was 1.8X/100 Hw without Palisades.

Projected PMR capacity factors are only half as sensitive to Palisades'nclusion through 1976 as they were through 1975.(See Appendix C for PMR trend equations with-out Palisades.)

Similarly, the BMR size impact through 1976, oaiftting the new Browns Ferry 82 data as well as Browns Ferry il, has only a 1.7X/100 Hw size effect (vs.1.6%/100 Hw through 1975, without Browns Ferry), but the confidence level is 92%(one-tailed test), compared to 87K previous-ly.BMRs admittedly do not show a conclusive size effect on capacity factor without the three Browns Ferry unit-years in 1975-76.Still, the BMR mean capacity factor is so low, at 55.6%through 1976, that the 62 JAG Ae allowance of even a small BMR size effect leads to sub-50K capacity fac-tor projections

'for new 1150-Hw BMRs.In short, the exclusion of Palisades and Bro~~s Ferry data, as sug-gested by the study critics, is not only not fully justified, but in ad-dition such exclusion does not have a confounding impact on the study's capacity factor projections, when 1976 data is added.-VP>>d~)>>dodd.>>~!,>>d d CRITICISM f2: The CZP ca cit factor trend ovorl sensl tive to removaI of several rl a~ele..nations ere said.to ho'fornfn unite from the'" Palisades, an 821-Hw PMR with a 1975 Ourn)laf:fve capacfty..factor of.23.2%(4 unit-years) and Browns Ferry I)1, a 1098-Hw BIN:with a 1975'orfe; year capacity factor of 14.t5, are large units which have suffered'..from

.'allegedly non-recurrent, non-size-related outages.Theft'fndlusfon.in the data base not only depresses the mean but>.accordfng to the industry,.-creates a spurious negative correlation.

between-capacity

'factor and unit size.--The suggestion that Palisades=and Browns Ferry 81 da'ta'be,'.excf'sed-fndcomputfng the projection mean is self-serving and without merit."""One-of-a-kind", outages are-a continuing fact'of life.ih-the'-nuclear.-

',: power'industry.

Their.putative uniqueness"in'each particu1ar'as'e has.'-not prevented other one-bf a-"kind outages from'causing low-cad pacfty fac-'-.61.<<q<<<<<<i<<<<~<~~-d~~

d<<d j~)>>dl<<-r'd<<d~>>d~'P i+QIP~dg)Q dtPP~X d, j 4)-y7d<<','~dd

<<d"r)<<>>~f~/vd dq dd d Q+QQgiv I Q@4 Q~g4~88/<<

I o~

f p t(, 70 6,%(5)67%('?54%(8)~51%(5)6)%(3)h flcC cs 30 I Hander o/plod fn otc cooror3<<39%(3)hW'i'2 3 6 5 6 7 3 9 (0 It)2"!3 Ycuh of commcrc4)he>vlcc Of plans hy t h'lg.1-Capactty factors ol nuctoar plants vs.ago ot plant, 1973-1974 t t ea h, I\I t~, I~~PA I"~V C I I.'~~~4 I I I'I~I I I~4~~~~I I I~N CJJ g 4 I I 4 I~I I~I I~~I~~~I~I~~0 I~I~N a ggC.~44 I j I'2I'I I I 4 I~~I'I~I I~~I~I I~I~I~I t~~~4 4~'t~'~~I I I I~I 4,'I 4 I I'I I I I~~I I I~~\'I I~4 I I~I~~I II~4~~I~~~I I 4 I~~4 4 4 I I 4'I'I~t~'~~4~I I~~I I~~, I~I I~I I 2 I~~I t I t I I I t I~~~~~~I t~I I~I~~~I~4 I~'I~~I I~I~~~I'I~~~4~~~CJ~4~~~I~~~4~~~~~I~I I I~~~~~~~~I I~I~~I 4~I 4 I I t*'5.~q,~~~~4 4~I I~I~I I~I I~4~~'I I~~~~~~~I~~I~I 4 I~~~I~~I~I~I~~4~~I~~~~I'I I~~~~~.F8~I~+~I'I~~~I 3("~~~I~~~C~~z-'7~~~~'I I~3/~~I~~~'I I~~

~e l 1 I I I 2~July 1977$6.1liam J.Dircks, Assistant Executive Director for Operations US Nuclear Regulatory Commission 3'lashington, DC 20555

Dear Nr.Dircks:

Concerning the letter that I sent you on5O.June 1977, on the subject of fuel efficiency, I made the same error in the letter several times consistently, so as not to affect the real proportions of electricity delivered.

Every time the term"MK'Ke" was used, the term should have read,"NNkWhe." A silly error, sorry.Sincerely, Ru sell.Iowes 7501 E.Hubbell Scottsdale, AZ 85257 C~o'+gal-&/)y~~Q*I