4 Project Structure, Cost, and Financing

4

Project Structure, Cost,and Financing

4c

Project Structure, Cost, and Financing

The only present way to transport natural gasacross ocean distances is to ship it as a liquid at— 260o F in specially insulated tankers. M e t h -.

ane, the principal constituent, is 600” timesdenser in liquid form than as a gas at room tem-perature, and this reduction in volume permitseconomical use of ships notwithstanding thecost of specialized liquefaction, revaporization,storage, and other terminal facilities.

Liquefied natural gas (LNG) projects are ex-pensive. The total capital required for a world-scale venture involving 1 billion cubic feet perday (Bcf/d) beginning in the early to mid- 1980s isnearly $.5 billion ( 1978 dollars). Approximately40 percent of this cost is applied to the gas pro-duction and liquefaction facilities in the export-ing nation, about 40 percent is needed for ships,and the balance of about 20 percent is for theimport terminal and revaporization facilities inthe United States. The cost of service, including

operating~ expenses and amortization of the ini-tial investment, for a typically structured proj-ect appears as a function of distance in figure12.

An LNG importer must pay in addition totransportation costs a return to the producingcountry for the wellhead value of the gas, andsupply contracts generally contain f.o.b. priceprovisions calculated to make imported gascompetitive with distillate fuels in the U.S. mar-

ket. However, escalation formulas in presentcontracts are such that delivered LNG pricesshould rise more sIowly than those of productsfrom foreign crude oil.

LNG projects

Figure 12.— Cost of Service as a Function ofDistance for a Typical LNG Import Project in the

Fifth Year of Operation (1990)

I I I Io 5,000 10,000 15,000

Distance (nautical miles)

SOURCE OTA, based on Jensen Associates data

Financial risk represents another element ofcost, and the public guarantees in part the com-mercial success of’ an LNG project through regu-lated retail prices designed to allow investors torecover portions of their cost notwithstandingsome kinds of failure or loss. on the other handthe risk of’ unilateral interruption of shipmentsby the supplier country is reduced by high capi-tal costs and a project structure that ties buyerand seller into a tight economic partnership.

Commercial I.NG trading began in 1964 withthe Algeria- United Kingdom project, involving().()4 trillion cubic feet (Tcf) of gas per year.Over the past 15 years, the international tradehas grown to 12 currently operating projects to-taling 1.75 Tcf/yr from six producer countries(table 24). Japan has the largest portion of pres-

ent imports (45 percent), f’ollowed by WesternEurope (29 percent) and the United States (26percent). However, the rate of future expansionin international LNG trade is uncertain. Shouldall the projects listed in table 24 materialize,worldwide trade in LNG would increase to 6.44Tcf/yr by the mid-1980’s, of which U.S. imports

69

59-406 2 - 80 - 6

70 ● The Future of Liquefied Natural Gas Imports

Table 24.–Operational LNG Projects, as of July 1,1979

Destination

ContractPurchasing Startup Volumesa

Origin Country Terminal companies date Tcf/year Remarks

OPERATINGAlgeriaArzew United Kingdom Canvey Is. British Gas Corp. 1964 0.04 Contract has been

Arzewextended

France Le Havre Gaz de France 1965 0.02Skikda France Fos Gaz de France 1972-73

1971-77 0.14Skikda United States Everett, Mass. Distrigas 1978 0.05Skikda Spain Barcelona Enagas 1976 0.55Arzew United States Cove Pt., Md. Columbia Gas, 1978 0.40

Savannah, Ga. Consolidated Gas,Southern Energy

AlaskaKenai Japan Negishi Tokyo Electric 1969 0.05 b

Tokyo Gas

BruneiLumut Japan Negishi Tokyo Electric 1972 0.26 b

Sodegaura Tokyo GasSemboku Osaka Gas

LibyaMarsa el Brega Spain Barcelona Catalana de Gas 1971 0.04Marsa el Brega Italy La Spezia Snare 1970 0.09

Abu DhabiDas Island Japan Sodegaura Tokyo Electric 1977 0.10b

IndonesiaBadak (Bontag) Japan Himeji Kansai Electric 1977 0.16b

Chita Chubu ElectricArun (Lhakseumawe) Japan Tobata Kyushu Electric 1978 0.22 b

Semboku Osaka GasNippon Steel

APPROVEDAlgeriaHassi R’mel Italy Sicily EN I 1981 0.44 Pipeline replaced

(gas pipeline) an LNG projectArzew Belgium Zeebrugge Distrigaz 1982 0.20 Terminal site

uncertainArzew France Montoir Gaz de France 1980 0.20Arzew/Skikda United States Lake Charles, La. Trunk line 1980 0.18Arzew/Skikda West Germany Wilhelmshaven Ruhrgas, Salzgitter, 1984 0.41

Netherlands Emshaven GasunieArzew/Skikda West Germany Wilhelmshaven Brigitta-Thyssen 1985 0.16

IndonesiaArun United States Pt. Conception Pacific Gas & Electric 1983 0.20 Approved Sept. 26,

So. California Gas 1979.

AlaskaCook Inlet United States Pt. Conception Pacific Gas & Electric ? 0.15 Approved Oct. 12,

So. California Gas 1979. Added re-serves needed

PROBABLEAustraliaDampier Japan Tokyo Tokyo Electric 1984-85 0.33

Tokyo Gas, etc.

Ch. 4—Project Structure, Cost, and Financing ● 71

Table 24.—Operational LNG Projects, as of July 1, 1979—continued

Destination

ContractPurchasing Startup volumesa


MalaysiaBintulu Japan Sodegaura Tokyo Electric 1983 0.31b

Tokyo Gas, etc.

IndonesiaBadak (exp.) Japan Various Chubu Electric 1983 0.16 b

Osaka GasKansaiToho Gas

POSSIBLE (active)NigeriaBonny United States/ Columbia, Mid 0.6

Europe Consolidated, 1980sSouthern, Mich-Wis,Trunkline and others

TrinidadPt. Lisas United States Gulf coast Tenneco 1984-85 0.18

Peoples

CanadaMelville Is. Canada/ St. Lawrence Southern Natural Gas 1982-83 0.09(Arctic Is.) United States

AustraliaDampier United States Pt. Conception So. California Gas late 0-0.15

Pacific Gas & Electric 1980’s

Cabo Negro United States Pt. Conception So. California Gas 1983-85 0.08Pacific Gas & Electric

IndonesiaA run (exp.) Japan Various 1985 0.12

POSSIBLE

Algerianot announced Sweden Wilhelmshaven Swedegas AB 1984-85 0.07 Trends in Swedish

energy policycast doubt onthis project

Gaz de Francenot announced France 0.18not announced Switzerland 0.000018not announced Austria Ferngas; OMV 0.07not announced Yugoslavia 0.07-0.11Arzew United States La Salle United, El Paso mid 0.40

El Paso 1980’s

Qatarnot announced Japan Tokyo Tokyo Electric mid 0.31b

Tokyo Gas 1980’sMitsubishiShell

Abu DhabiRubais Japan C. Itoh & Co. mid 0.25 b

1980’s

Colombia United States 0.05U.S.S.R.Yakutsk United States Tokyo Gas 0.75

Japan Tokyo ElectricEl PasoOccidental

—


Table 24.—Operational LNG Projects, as of July 1, 1979—continued

Destination

ContractPurchasing Startup volumesa


Murmansk United States 0.75Europe

United KingdomNorth Sea United Kingdom 0.75 Floating barge

liquefactionplant.

Iran Japan 0.13Thailand JapanChina JapanNew Zealand Japan Maui gas, Mobil

has proposed anautomotive fuelproject

aAt I,1020 Btu/cf Normally contract volumes are given f o b the Iilquefaction plantblndlcates c. i. f. volumes, I e , delivered.

SOURCE Jensen Associates, Inc.

would account for about 36 percent. Not all ofthese projects will come to fruition, however,and most past projections regarding the futureof LNG trade have overestimated the rate ofgrowth. The possible level of LNG imports isparticularly uncertain in the U.S. market,where Government policy regarding LNG im-ports has been difficult to predict. The Depart-ment of Energy (DOE), the Federal Energy Regu-latory Commission (FERC), and State Public Util-ities Commissions decide on all aspects of indi-vidual projects case-by-case in regulatory pro-ceedings that take years, Given the present un-certainties, a more reasonable expectationwould be that worldwide trade in LNG willreach 4.19 Tcf/yr by 1985, of which 46 percentwill move to Western Europe, 34 percent to Ja-pan, and 20 percent to the United States.

A baseload LNG project is a complex andhighly capital-intensive venture, consisting of’three primary segments (figure 13):

1. liquefaction, storage, and loading facilitiesin the producing country;

2. transportation facilities (cryogenic tank-ers); and

3. terminal and revaporization facilities in thereceiving country.

Total capital investment of a 1 Bcf/d projectcan exceed $5 billion (1978 dollars). The cost

varies with such factors as the gas-gatheringsystem, shipping distance, and new deliverypipelines required. The cost of liquefaction andrelated facilities in the producing country canaccount for as much as 50 percent of overallproject costs. 1

What follows is a more detailed description ofthe physical and cost structure in LNG importprojects, including the price policies of the ex-porting countries. Two LNG projects are usedfor the purpose of illustration: Pac Indonesiaand Algeria 11. Although only one of them hasreceived final U.S. Government approval as ofthis writing, * the projects are good examples,because their costs and pricing provisions arerecent and represent current LNG trade.

Algeria II

The proposed project was based on an Octo-ber 28, 1975, contract, as amended, betweenSonatrach (Societe Nationale pour la recherche,

1 FOI” (l(’l’(’lol)llll’ 111 01 l~l)i(>al 1,N(; (os1 (ISI In):i[(}s, S(I(I K N’.

l)] N’iiPOli, “l;still]altn, q (;os1s lo]” 11;1s1’-10;1({ I ,S(; l]liilll S,” oi/ iil?d (;;1S

.h)llrnd, N()\ 17, 1 97.;.“1’llt’ ,,llg(>l’iii II 1)1’oj(’(’l \\ il~ (’oll(lillolliill~’” ill)l)l’()\ (’(l 1)}’ 111(1 };l’(:

I; F;K(’ il(illlillihl l’iilil (’ IiI\\’ j(l(lg(’ 011 ()(’t. 25, 1977 lio\i (’\’(’l’, 1111( 1(’1’

[11(’ l)(>j)ill’l Illt’111 ot liIl(’1’~J’ ol’~iilli~iltloll” ,\(’l ( ’ l ’ t ( ’ ( ’ l l\ ( ’ ()(’1 1, I !)77,

I 11]1)01’1 jurih(l 1(’1 i o n \f ilS t I’illlst (’1’1’(’(1 10 I X )I;’s l;ll(’1’gJ’ 1{(’glllii 101’}”

.\(!ll)l])lstl’<illotl IFIR l] 1)01:, t;,R,1 IIIA (IIWK1 III(’ Illiliiil (I(l(isio[l III IIS

of)l[lioll” N() 4 ot I J(I(. 21, 1 !)78 P(lrn)lssioll 101” l’(’11(’iil’ill~ 11:1s 1)(’[}11

[’(’(’(’111 IJ’ ~1’iill[(’(1

Ch. 4—Project Structure, Cost j and Financing ● 73

Figure 13.— Major Segments of an LNG Import Project

SOURCE: Jensen Associates, Inc.

1a production, le transport, la transformation etla commercialization des hydrocarbons) and ElPaso Atlantic Company (a subsidiary of ElPaso). 2 It provided for the sale of LNG contain-ing 410,625 billion Btu annually, for a term of 20years. This amount is equivalent to approxi-mately 1 Bcf/d of natural gas at 1,125 Btu/cf.

The gas was to be produced by Sonatrach, theAlgerian State oil and gas company, in the Sa-hara, pipelined 315 miles to the Mediterraneancoast, and there liquefied, stored, and loadedaboard LNG tankers. El Paso Atlantic, whichwould acquire the title to the LNG at the tank-er’s receiving flange, would arrange for thetransportation by a fleet of 12 cryogenic tankersto an import terminal and regasification plant(the La Salle terminal) located near Port O’Con-nor, Tex. Six of the ships would be provided bySonatrach and six by Atlantic. As each vessel en-tered the international waters off the coast ofA]geria, the title to the LNG would pass to ElPaso Eastern, the legal importer. The La Salle

‘The El Paso Companies involved in the project, and their gen-

ealogy, are as follows:

The El Paso Company

El Paso LNG Company El Paso Natural Gas Company

El Paso Eastern Company ("Eastern") [El Paso Natural)El Paso LNG Terminal Company

(“Terminal”)

El Paso Atlantic Company [“Atlantic”)

SOIIR(’L /nllla/ Dem$mn, 1 jwn ,[email protected] 10 In]lxmf f,,\’(; f’rom Al,qfwa, F’i.R( 01. I 2.5,1979, Docket N(M (.11 77.330, N <i] p 4

terminal facilities would be built and operatedby the El Paso LNG Terminal Company whichreceives, stores, and revaporizes the LNG. Atthe outlet of the terminal, the gas would be soldby El Paso Eastern: 65 percent to the El PasoNatural Gas and 35 percent to the United LNGCompany. The entire quantity would be pipe-Iined to the United Gas Pipeline Company’s ex-isting mainline facilities near Victoria, Tex.There, United LNG’s 35 percent of the gaswould be sold to its parent, United Gas Pipeline,which serves other major pipelines that delivergas throughout the area east of the Mississippi.The remaining 65 percent of the gas would betransported via a new 432-mile-long pipeline tobe built by El Paso Natural to its Waha treatingplant located in Reeves County, Tex., where itwould enter the present El Paso system servingthe Southwest and California.

In 1977, at the time of the participants’ initialapplication to the Federal Power Commission(FPC) for import authorization, the total capitalcosts of the project were estimated as follows:*

● $2,300 million for gas welIs, pipeline, andliquefaction facilities in Algeria (including$391 million for interest on funds used dur-ing construction);

● In 1975-76 dollars.


● $1,752 million for 12 vessels and shoresidefacilities required for ocean freight; and

● $719 million for receiving terminal, regasi-fication plant, and new pipelines in Texas.

Pac Indonesia

Two gas utilities in California—Pacific Light-ing Corporation (PLC) and Pacific Gas and Elec-tric Company (PG&E)—have formed a partner-ship to import LNG from Indonesia through twosubsidiaries. The first subsidiary, Pac Indonesia,has entered into a contract with Pertamina, theIndonesian Government-owned oil and gas com-pany, for the purchase of 226,194 billion Btu an-nually (approximately 550 MMcf/d) for a periodof 20 years. *

The gas for the project would be produced inthe Arun field of Northwest Sumatra by MobilOil Indonesia, Inc., under a production-sharingcontract with Pertamina. From the field, the gaswill be transported via a 20-mile pipeline to theliquefaction plant, which will be owned and fi-nanced by Pertamina.

Pac Indonesa has entered into contracts forthe hire of nine cryogenic tankers to transportthe purchased LNG from North Sumatra to Cal-ifornia. Three of the vessels have already beencompleted in foreign shipyards and plans callfor the remaining six to be constructed in theUnited States.

The LNG would be delivered to a proposed re-ceiving terminal to be constructed by WesternLNG Terminal Associates, the second subsidi-ary, near Point Conception, Calif. After storageand revaporization, the gas will be transportedvia a new 112-mile pipeline to the transmissionsystems of PLC and PG&E, which will jointlyown the pipeline. Pac Indonesia will sell half ofthe gas to Southern California Gas (So Cal), awholly owned subsidiary of PLC, and the otherhalf to PG&E. The two utilities combined com-prise the transportation and marketing mecha-

“rhe original contract between Perusahaan Pertamban&in Min-

yak Dan Gas Bumi Negara (Pertarnina) and Pac Indonesia’s prede-cessor—Pacific Lighting International, S, A.—was signed in Septwn-

ber 1973, Since then, it has been amended three times in regard to

its priring provisions. ‘1’he last amendment was introduced in Ju ly1978.

nism that handles virtually all natural gas con-sumption in California.

Based on 1976-77 cost estimates, the capitalexpenditures for the project are as follows:

●

●

●

$869 million for the pipeline, liquefactionplant, and related facilities in Indonesia,(including an estimated $164 million for in-terest during construction but not the costof developing the Arum gasfield);$1,230 million required for nine charteredtankers, including $930 million for six ves-sels to be built in the United States (at $155million per ship);$436 million allocated for the receiving ter-minal and pipelines in California. These fa-cilities, estimated to cost a total of $749 mil-lion, are to be shared by Pac Indonesia andPac Alaska. On the basis of the contractedthroughputs, the cost allocated to Pac Indo-nesia would be just over 58 percent of thetotal.

Pricing policies of exporting countries

As a consequence of large crude oil price in-creases in 1973-74, the LNG projects negotiatedor renegotiated after 1974 contain fuel-relatedescalation clauses applicable to their base f.o.b.ship’s rail prices, the purpose of which are to es-tablish parity between LNG and alternativefuels. Minimum (floor) price levels designed toremove the producing country’s investment, orto assure the timely repayment of project-re-lated debt, have also become standard contrac-tual provisions. In addition, the pricing formu-las usually contain safeguards against currencyfluctuations. Sonatrach has adopted a fairly uni-form f.o.b. pricing policy for all of its recentcontracts—U. S. and European alike. A review ofthe major price provisions in the Algeria II andPac Indonesia contracts provides a good indica-tion of a LNG pricing mechanism that typifies allrecent LNG trades.3

.lFor a more detailed discussion of I.N(; pricing mechanisms, see

“Economic Considerations and Operating History of Base-Load

LN(; Projects, ” Philip J. Anderson and F;dward J. Daniels, lnstitutt~

of Gas “technology, December 1977.


The procedural history of pricing clauses ne-gotiated and approved in the Pertamina-Pac In-donesia contract illustrates the evolution of pol-icy involved in f.o.b. pricing. Under the originalSeptember 6, 1973, Pertamina contract, theprice to be paid by Pac Indonesia’s predecessorwould be $0.63/million Btu (MMBtu) plus 2-per-cent annual escalations, adjusted by a currencyreevaluation factor and subject to certain floorand ceiling levels. The Indonesian Governmentdid not, however, approve the contract on theground that the price formula which containeda fixed escalator would not reflect the develop-ment of world energy prices in general and, inparticular, was not linked to the price of In-donesian crude oil. Consequently, the firstamendment issued January 9, 1975, establisheda new f.o.b. base price of $l.25/MMBtu—ap-proximately double the prior price—and deletedthe fixed 2-percent-per-year price escalator. Anew escalation formula reflected equallychanges in Indonesian crude oil export pricesand U.S. energy prices as measured by theBureau of Labor Statistics wholesale index forfuels. The renegotiated formula no longer con-tained a floor or a ceiling, so it offered no pro-tection to either party against potentially widefluctuations in LNG price through the operationof the escalation clause. The possibility of a fallin crude oil prices presumably led to the mini-mum bill provision, which assured Pertamina’slenders that the price of LNG would be at leastsufficient to service Pertamina’s debt and tomeet operating and maintenance expenses (sec-ond amendment, issued October 28, 1975).

Although the FPC administrative law judgeconditionally approved the proposed projectand its pricing provisions, one of FPC’S succes-sor agencies, DOE’s Economic Regulatory Ad-ministration (ERA), did not allow the automaticflow through of cost increases under the priceescalator clause, charging that the provisionwas tied too directly to future movements inOPEC-administered prices, and that the U.S.fuels index would be influenced by future do-mestic energy pricing policy and by the price ofthe import itself; thus creating a significant self-compounding effect.4 This rejection of the esca-

later led to yet another price amendment, is-sued July 28, 1978, and approved by DOE/ERAshortly thereafter.

Under the renegotiated escalation clause, theIndonesian half of the escalator will still be tiedto Indonesian crude oil export price, but withthe added constraint of a 15-percent absolutelimit on annual fluctuations in that price. Anyadjustment above the 15-percent absolute limitor below the floor can be carried forward untilit can be applied. The U.S. half of the escalatorwas changed to substitute the broader basedBureau of Labor Statistics “all commodities” in-dex for the former fuels-related index.

The pricing formula, as finally approved, isshown in figure 14. The calculated contractsales price is $1.25/MMBtu multiplied by theequally weighted changes in the Indonesiancrude price (subject to a limit on annual fluctua-tions) and in the U.S. wholesale index for allcommodities. A contract sales price is then mul-tiplied by a currency reevaluation factor to ar-rive at the billing prices. *

If at any time during the debt amortizationperiod, the calculated contract sales priceshould be lower than the minimum price calcu-lated by Pertamina, the latter will be the billingprices

The Pertamina-Pac Indonesia contract in-cludes a “most favored nation clause” underwhich Pac Indonesia would be entitled to a con-tract sales price for LNG no higher, on an f.o.b.equivalent basis, than that paid by any other im-porter under any other contract with Pertaminain existence as of January 9, 1975. Otherwise,the contract does not provide for future pricereviews.

The Algerian pricing system has a twofoldpurpose: 1) to ensure that imported gas is com-

‘ iI()\%(?\(Ir, operation of the (wrrwq factor cannot reduce thebilling pri(>e t)dot~ M’tuit i t was o n the difte ot’ f i r s t cteli\wries, n o r

i n c r e a s e it morf? than M perrf>nt ahm’e ~hf~ r)t hf!rjt’isf~ applicat)lv

price in any gilfm c a l e n d a r qLmrler.

5’[’hf> FIY adrninist rat i~re la~$; juc!gf> t$’ho con(iit iona]lv ir}]prolred

the Piic In[lonwia projecf interpretcc~ (hc min imum h i ; l sut)jc(.t [0

all the l~rol isions o] [hc sales contract. ‘1’}lLIS, Pa{’ lndonesia ki Ould

n o t I)f’ rfquird to pa)’ for qutintilies rrol ddilwd ‘(whether l))’rf’irson of PfJrtaminii’s fii Lilt or force m~i jeuix’ or :issim iliit(’[i (ii’-(mmslti ilws uccuriilg i n an-v ])ii rt of t 11[’ fii(’il it if’s, in(>lud ing t Ilf’

s h i p s o r tf?rminal. ” (Init iii] Ilfx. ision, p 62)

76 . The Future of Liquefied Natural Gas Imports

Pac Indonesia

Figure 14.—Pricing Provisions of Pac Indonesia and Algeria II Import Projects(U.S. dollars per million Btu, gross heating value, loaded f.o.b.)

Algeria II

Contract ("invoice") price

Calculated quarterly. Calculated semiannually.

F1

= price for No. 6 fuel oil, low pour, max. sulfur of0.30%, delivered New York harbor.

F lo = $13.505.

B =

cl =

C2 =

on the date of initial deliveries for eachof the currencies.the arithmetic average of the commercialrates of exchange on the applicable datesin each quarter for each of the currencies.

B = 1 until its absolute value changes at leastby 0.1 O/O. Thereafter new value for B usedonly if it differs from old by 0.1 0/0 or more.

Maximum B = 1.25.

M I : , , * , ;.“’

. , - ,

During its debt amortization period Pertamina willcalculate a price sufficient to meet:

1)

2)

3)

E = arithmetic average of the results obtained byapplying the formula:R~– 1 to each of 6 currencies.

A = average commercial exchange rate for eachcurrency during July 1975.

B = average commercial exchange rate for eachcurrency as measured by average purchase andsales rates for telegraphic transfer for eachbusiness day of preceding month.

E = O until its value increases by at least 0.1%as compared to O. Thereafter new value for E usedonly if it differs from old value by 0.1 0/0 or more.

M P 1 = recalculated minimum price.

x = actual capital costs incurred by Sonatrach(in millions of dollars),

Y = actual operating costs of Sonatrach during thefirst year of operations (in millions of dollars).


adjust the minimum price only upwards, andwith no upper limit.

Since oil prices are likely to continue rising,the contract (“invoice”) price rather than theminimum price will probably determine Sona-trach’s billing price. Recently, Sonatrach and ElPaso’s subsidiary have renegotiated the invoiceprice formula in a 1969 contract which under-lies the Algeria I project. * Under the 1969 con-tract, the current price for LNG f.,o.b. Algeriawould amount to some $0.363 /MMBtu. In ra-tionalizing the price renegotiation, Sonatrachhas observed that “in the decade since signing ofthe contract, the capital cost and operating costsof the project have increased substantially, andthat, as a result, Sonatrach is suffering a huge fi-nancial burden while providing the cheapest in-cremental source of natural gas to the UnitedStates,”7 The renegotiated base price will be$1.75/MMBtu, effective as of July 1, 1979. Aseries of discounts will be applied to this priceranging from $0.60/MMBtu for the remainderof 1979 and then decreasing by $0. 10/MMBtu in-crements until mid-1983. The price escalator istied to No. 2 and No. 6 fuel oils as described forthe Algeria 11 project.

Unlike the Pertamina contract, the Sonatrachagreement provides for the regular review ofthe contract sales price. The parties are ex-pected to meet during the first year after regu-lar delivery begins, and every 4 years there-after, to ascertain whether the prevailing priceof the gas resulting from this project is still com-petitive in the U.S. energy markets. Further-more, either party may request a meeting at anytime if the particular indices selected to reflectfuel oil prices in the U.S. market fail to do soadequately.

This important price provision may result insignificantly higher f.o.b. prices to Sonatrachthan would otherwise prevail without renegoti-ations. The reason is the disparity, which is


likely to develop over time, between the U.S.tariff imposed on the regasified LNG and theprice of oil in U.S. markets (figure 15). In light of

Figure 15.—Comparison of the Forecast U.S. Tariffon Regasified LNG in the Algeria II Project With

the Delivered Price of Fuel Oil*●

designed. Operating costs are subject to in-flation, but they constitute a small portionof the tariff. For simplicity, the graph re-flects the assumption that the shipping andterminating components of the tariff willremain fixed.

Because the f.o.b. price of LNG grows more

7.0

6.0 I Delivered price of‘— oil in U.S. market

U.S. tariff A—

on regasi f ied LNG A

f.o.b. price of LNG

‘ - without renegotiat ions

2.01

2.

Light Arabian marker crudeprice increases by 420/.In 1979; by 12°/0 annually1979-82; by 170/. per yearuntil 1990.

U.S. inflation rate equals 70/.

1 I I I I I I I I I !1978 79 80 81 82 83 84 85 86 87 88 89 1990

Year

. Every 4 years a portion of the difference between the delivered priceof oil and the U.S. tariff on regasified LNG is liable to Sonatrach’sclaims through the operation of the price renegotiation clause.


the price review provision, Sonatrach mayclaim this potential price differential for its ownbenefit. The price disparity–represented on thegraph by the darker tone–will occur for the fol-lowing reasons:

● The LNG (f.o.b. ) price component of thetariff grows in the same proportion as theoil price, but since this rate of growth ap-plies to a smaller base, the dollar differencebetween the oil price and the f.o.b. pricefor LNG increases in time.

● The shipping and terminating componentsof the tariff consist largely of capitalcharges, which are either fixed or declin-ing in time –depending on how the tariff is

slowly in terms of-absolute dollars than theoil price, while the shipping and terminat-ing costs are fixed, a gap develops andgrows between the price of oil and the tar-iff for the regasified LNG.

Revisions of f.o.b. price may well serve as avehicle for liquidating such disparity by addingthe price differential to Sonatrach’s f.o.b. pricefor LNG. It should also be noted that by keepingthe price of the gas competitive with that of oil,the price revision clause assures the marketabil-ity of Algerian gas in the United States. For in-stance, should the regasified LNG become moreexpensive than reference New York harbor fueloils, the price renegotiation clause may be in-voked to bring the price of Algerian gas down tothe competitive level.

Conditional approval of the Algeria 11 projectby the FPC/FERC administrative law judge onOctober 25, 1977, was subsequently reversed byDOE/ERA. Much of the ERA criticism of Sona-trach’s price provisions echoed its earlier objec-tion to the Pac Indonesia pricing mechanismprior to the issuance of the last amendment.ERA objected mostly to the fact that the Sona-trach price escalator is entirely linked to futurechanges in OPEC-determined prices of premiumpetroleum products. The formula was foundlacking safeguards against extreme oil price in-creases, since it imposed no limits on the annualprice fluctuations. ERA also criticized the use ofNo. 2 and No. 6 posted prices rather than theweighted average of the actual transactionprices (the latter practice is proposed for thePac Indonesia project). DOE/ERA, however, in-dicated that its approval of the Pac Indonesiaproject did not create a precedent for subse-quent decisions. In other words, should Sona-trach adopt exactly the same price provisions asPertamina, the project would not necessarily beapproved on those grounds alone.

Ch. 4—Pro/ect Structure, Cost, and Financing 79

Figure 16 depicts the forecast f.o.b. prices de-rived from Sonatrach’s and Pertamina’s formu-las assuming no price renegotiations. Under the

Figure 16.—Forecast f.o.b. Prices Paid for LNGin Pac Indonesia and Algeria II Projects*

I I 1 I 1 I 1 i. I I I1978 79 80 81 82 83 84 85 86 87 88 891

Year‘Assumptions:

1. Light Arabian marker crude price increases by 420/. in 1979; by 12°/0annually 1979-82; by 17°A per year until 1990.

2 U.S. inflation rate equals 70/. annually.

3. Sonatrach’s formula IS not periodically revised.

● “The two curves for Pac Indonesia show f.o.b. price calculated with andwithout the 15 percent ceiling on the annual fluctuations In the valueof the Indonesia half of the price escalator.


listed assumptions, Algeria 11 prices would in-crease considerably faster than Pac Indonesia’s.This difference is due to two factors:

1. the expected rate of growth in the price ofimported fuel oil is well above the pre-sumed U.S. inflation rate; and

2. the annual ceiling on the Indonesian crudeprice increases limits the impact of theproject price hikes.

Producing country facilities andrelated costs

PAC INDONESIA

The Pac Indonesia project entails the follow-ing operations in Indonesia:8

1.

2.

Production and gathering by Mobil Oil In-donesia of natural gas from the Arun fieldin North Sumatra and transportation of thegas via pipeline to Pertamina’s liquefactionplant and marine terminal on the northcoast of Sumatra.Liquefaction, storage, and delivery of theLNG by Pertamina to the LNG vessels char-tered by the Pac Indonesia LNG company atPertamina’s marine terminal.

The source of the gas is specified in the Per-tamina contract as Contract Area “B” in theAceh Province, * which contains the inlandArun gas condensate field discovered by MobilOil Indonesia in late 1971. Arun’s proven re-serves consist of an estimated 13 Tcf of nonas-sociated gas. For an LNG import project, nonas-sociated gas is preferable because the availabil-ity and stability of its supply is not adversely af-fected by potential interruptions and otherproblems in crude oil production.** Pertaminahas contracted to sell LNG produced from Arunnot only to Pac Indonesia by also to a group offive Japanese purchasers, who are scheduled toreceive a slightly greater average daily volume.

The field is being developed by Mobil Indo-nesia, a wholly owned subsidiary of Mobil OilCorporation, under a production-sharing con-tract with Pertamina. Eventually, 64 wells (withan average depth of 11,483 ft) will be needed tomaintain an adequate gas supply for both Pac

‘For description, see Initial Derision on Importation of LiquefiedNatural Gas From Indonesia, FPC, July .22, 1977, Docket NO.. CP

74-10 et al.

● Other producer countries, for instance Algeria, do not dedicatespecific gas reserves to the fulfillment of individual contracts All

Algerian gas reserves stand twhind all of its contracts.

* *For instance, Libyan gas is normally f’ound associated with

crude oil and therefore gas availability depends to a great extent

on crude oil production. Conservation policies in I.ihyan crude oil

production will l imit the quantities of gas available for l iquefac-

tion.

80 The Future of Liquefied Natural Gas Imports

Indonesia and Japanese contracts. Due to unu-sually high reservoir pressure and temperature,each wellhead has to be equipped with speciallydesigned piping and valves to control the gasstream.

From the field, the gas is transported to theArun liquefaction plant at the north coast of Su-matra via a 42-inch-diameter, 20-mile-long pipe-line with a design capacity of 1,777 MMcf/d–sufficient to transport the quantities of gas toservice both the Japanese and the Pac Indonesiacontracts. * As shown in table 25, the capital costof the pipeline attributable to Pac Indonesia(half of the total) is $13 million.

Table 25.—Estimated Capital Costs of lndonesian-Based LNG Facilities for Pacific Indonesia Project’

(millions of dollars)b

Amount TotalPipeline. . . . . . . . . . . . . . . . . . . . . . . . . . . .Plant facilities

Gas treating and liquefaction(3 trains) . . . . . . . . . . . . . . . . . . . . . . .

LNG storage and loading . . . . . . . . . . .Plant utilities . . . . . . . . . . . . . . . . . . . . .Site development, buildings,

miscellaneous . . . . . . . . . . . . . . . . . .Contractor’s home office costs. . . . . .

Supporting facilitiesHousing . . . . . . . . . . . . . . . . . . . . . . . . .Communications facilities. . . . . . . . . .Other. . . . . . . . . . . . . . . . . . . . . . . . . . . .

IntangiblesProject management. . . . . . . . . . . . . . .Pre-startup and training costs . . . . . . .Other (land, insurance, taxes,

royalties, misc.) . . . . . . . . . . . . . . . . .Contingencies . . . . . . . . . . . . . . . . . . . . . .

Subtotal . . . . . . . . . . . . . . . . . . . . . . . . .Interest on funds used during

construction:. . . . . . . . . . . . . . . . . . . . .

Total . . . . . . . . . . . . . . . . . . . . . . . . . .

$ 13

2028159

8872 502

4967 62

2518

35 7850

705

164

$869

aThese cost estimates include.-The construction of liquefaction trains 4,5, and 6, assuming that procurementwill take place in the world market and that mechanical completion of the 4th,5th, and 6th trains will fake place in May, August, and November 1981 respec-tively.

-One-hlalf of the cost associated with the “common” facilities required to serv-ice all SIX liquefaction trains. Interest during construction IS not included.

bresumably 1976 dollars.cEstlmated by Jensen Associates, lnc.

SOURCE: Testimony of President.Director of Pertamlna, Piet Harjono, beforethe Federal Power Commlsslon, Feb. 25, 1977. Exhibit No. 175,FPC Docket No CP-74-160.

The liquefaction plant converts the naturalgas received from the pipeline into a liquid suit-able for storage. A liquefaction facility consistsof three main sections.

1.

2.

3.

Gas preparation section-Any constituents,such as water vapor, which freeze at lique-faction temperatures and thereby plug thecryogenic equipment, must be removed.Removal of hydrogen sulfide is also re-quired to meet LNG product specifications.Liquefaction section—Mechanical equip-ment refrigerates the gas in order to lique-fy it. At atmospheric pressure, the gas be-comes a liquid at – 260° F and its volumediminishes by a factor of 600.Storage and loading section—Insulated tank-ers retain the natural gas as a liquid at at-mospheric pressure, and the loading sys-tem transfers the product from land-basedstorage to oceangoing tankers.

Approximately 3 years are required for thecomplete design and construction of a large liq-uefaction plant.

The liquefaction facilities proposed for thePac Indonesia project (the Arun plant) representequipment, processes, and costs that are typicalfor contemporary large-scale LNG plants. Theoverall Arun plant will include six liquefactiontrains (three for the Japanese project and threefor Pac Indonesia) together with feed gas pre-treatment, refrigerant preparation and storage,LNG loading, and required offsite and utilityfacilities.9 The first three liquefaction trainshave already been completed and, since August1978, are serving Pertamina’s obligations to theJapanese clients. The design and construction ofthe first three trains anticipated the projectedsix-train operation in terms of sizing, location,and utilities layout. This sharing of facilities


Photo credit El Paso Co

Frost forms at the flange and on the articulating arm ascold LNG flows onto an LNG tanker at the

loading terminal

provides for convenience in operation and sav-ings in capital costs.

Each train is designed to produce LNG equiva-lent to 200 MMcf’/d* in 341 days Of annual op-eration. Three trains would therefore produce102 percent of the annual quantity contractedfor by Pac Indonesia. Indeed, the Indonesianplants that serve Japanese contracts (Arun aswell as the somewhat older Badak plant) haveconsistently produced well in excess of theirdesign capacity. Reliability of production isenhanced by the fact that both gas-processingand liquefaction trains are arranged in parallelso that the failure of any one component will

not result in a plant shutdown. Table 25 indi-cates that liquefaction equipment representsthe greatest portion of direct costs-about wpercent. Pertamina estimateci the cost of one liq-uefaction train to be constructed for Pac Indo-nesia at $67 million, assuming that procurementwould take place in the world market and thatall three trains would be completed by the endof 1981. *

LNG will be stored in four double-walled insu-lated tanks of 125,000 m3 each. The combinedcapacity of the four tanks equals 8.5 days ful]production of’ the six-train plant. The loadingsystem utilizes four pumps (with a fifth asa spare), which drain LNG from the tanksthrough two insulated pipes. The pipes termi-nate in loading arms that accommodate the rela-tive movement of the ship and the pier. The sys-tem is capable of loading a 125,000 m3 ship in 12hours at either of two berths. The total cost ofLNG storage and loading facilities is $162 mil-Iion, half of which constitutes Pac Indonesia’sshare.

ALGERIA II

The Algeria 11 project111 provides for daily de-livery of approximately 1 Bcf/d, a volume closeto the combined Pac Indonesia and Japanesecontractual amounts. Liquefaction, storage, andloading facilities proposed for the Algeria 11project are very similar to the ones describedfor the Arun plant. The six-train liquefaction fa-cility at Arzew will use the same air productsand chemicals (APCI) liquefaction process* * asin the Indonesia project. Plants are similarly ar-ranged m parallel independent equipmenttrains. However, the facilities are designed toproduce 105 percent of requireded yearly quanti-ties in 330 days, thus, in theory, providing agreater allowance for downtime than the Arunplant (102 percent in 341 days). On the otherhand, the Arzew plant will have relatively lessstorage space than the one at A run. Arzew will


have three storage tanks, each with capacity of100,000 m3, to accommodate its 1,000 MMcf/dproduction, compared to Arun’s four 125,000m 3 tanks for the combined Pac Indonesia-Japan-ese production of 1,131 MMcf/d. Loading facili-ties are similar in both countries. Another com-mon feature is sharing of equipment amongprojects. The six proposed trains for Algeria 11will share certain supporting facilities—such asthe cooling water system, steam system, and ad-ministration—with those already serving the Al-geria I project, and the marine terminal will alsoserve other future projects.

As can be seen from table 26, the estimatedcapital cost of liquefaction and supporting facili-

Table 26.—Capital Costs per Million Btu ofDaily Contractual Quantity(1976 dollars/million Btu/day)

Pac Indonesia Algeria II

Pipeline from gasfield toliquefaction plant . . . . . . . . . . . . $ 21 $ 360

Liquefaction, storage, and loading 1,117 1,134

Subtotal. . . . . . . . . . . . . . . . . . . . 1,138 1,494Estimated interest on funds used

during construction . . . . . . . . . . 265 348

Total. . . . . . . . . . . . . . . . . . . . . $1,403 $1,842


ties per million Btu of contracted daily produc-tion is comparable in both projects, reflectingsimilar processes and equipment. Sonatrach es-timates the total capital cost of its Arzew plantat $1,276 million.

The most significant difference in the costs ofthe two projects lies in the respective field andpipeline systems. The Hassi R’Mel field, whichwill Supply gas for the Algeria 11 project, * re-quires only 22 wells with an average depth of7,054 ft to supply the contract quantity. In com-parison the Indonesian Arun field requires 64wells with an average depth of 11,483 ft, plusspecial stream control equipment, to produce asimilar amount of gas. These factors influenceproduction costs, since, for example, drillingcosts rise almost exponentially with well depth.Sonatrach has estimated that its field facilitiesfor Algeria 11 would cost $228 million.

While the Pac Indonesia project requires onlya 20-mile, 42-inch pipeline between the field andthe liquefaction plant, the cost of which wouldbe shared with the Japanese purchasers, Sona-trach plans to construct a 315-mile-long, 40-inch-diameter pipeline exclusively for the Alge-ria II project between Hassi R’Mel and the lique-faction plant at Arzew. Gas turbines at five com-pressor stations will maintain the pressure andflow. The estimated cost of the pipeline is $405million, and as shown in table 26 the capital costit represents per million Btu of daily contractedquantity is 17 times higher in the Algeria II proj-ect than in Pac Indonesia, reelecting the differ-ence in the mileage. * Sonatrach estimates thetotal construction funds to be $2,300 million,and the annual operating cost at $60 million(1976 prices).

Transportation facilities—cryogenictankers

Although they resemble conventional tankersin many ways, LNG carriers are highly special-ized, with designs strongly influenced by theunique characteristics of LNG—especially itslow density, cryogenic temperature, and flam-mability. 12 The principal feature is extensive in-sulation of the tanks to minimize vaporizationen route and to protect parts of the ship’s struc-ture that would be damaged by extreme cold.

For the actual arrangements of LNG shipping,several alternatives are available. An importer,or exporter, may own the vessels or operatethem through bare-boat charters, contracts ofaffreightment, time charters, or leverage leasearrangements. The proposed shipping arrange-ments for Algeria 11 and Pac Indonesia illustratetwo of these alternatives.

The Algeria 11 fleet would consist of 12 tank-ers, each with cargo capacity of 125,000 m3. Sixof the vessels would be furnished by Sonatrach,


e -.

Photo credjt Marty Saccone

Modern LNG tankers typically carry 125,000 m3 of liquefied gas

the other six by El Paso Atlantic, presumably the fleet will transport 143 loads of LNG annual-through indivual shipowners. IS

Each carrier will have an average servicespeed of 18.5 knots and will be capable of com-pleting the round trip voyage of about 10, I50nautical miles between Arzew and the La Salleterminal in an average of 28.4 days. With eachship operating from 332 to 333 days per year,

ly, and a ship will arrive at the La Salle terminalapproximately every 2.5 days. The energy deliv-ered by the LNG carrier fleet for use in theUnited States will represent about 95 percent ofthe quantity loaded at the Arzew terminal. Thesmall amount of vapor that boils off during thetrip is consumed as fuel in the ship’s boilers.

In addition to the double hull, other safety fea-tures of the carriers include a computerized col-lision avoidance system, bow thruster, lead-de-tection systems, dry-chemical and water fire-fighting systems, two complete navigational ra-

84 . The Future of Liquefied Natural Gas imports

dar systems, and five separate communicationsvstems. 14.

The estimated yard cost per vessel, con-structed in a foreign shipyard—or in a U.S. ship-yard, after construction differential subsidy*—would be about $106.5 million at 1976 prices.Other direct and indirect capital costs relatingto the vessels (see table 27) would bring the esti-mated capital investment per ship to $142.6 mil-lion. Shore-based facilities for all] 12 vesselswould be supplied by Atlantic at an estimatedcost of $40 million. Thus, assuming that thesame capital cost is required for Atlantic’s andSonatrach's vessels ($142.6 million per tanker),the aggregate investment by Atlantic would be$896 million, and $856 million by Sonatrach.The total estimated capital cost for the Algeria IItanker fleet would therefore amount to $1,752million, or $l)639/MMBtu/d.

Operating costs of LNG vessels are a functionof trip distance. For Algeria 11, the total fleet op-erating expenses per year have been estimatedat $72.5 million (1976 prices). Atlantic's oper-ating cost—for three foreign and three domesticvessels--would amount to. about $38 million an-nually (see table 28). The corresponding ex-penses for Sonatrach’s vessels are expected tobe the same as those estimated for Atlantic’s for-eign vessels, 15 and therefore would tot al about$34.5 million. The total operating costs amountto $(). 19/MIMBtu delivered in the Algeria 11 proj-ect.

The Pac Indonesia project involves a differentshipping arrangement. To transport the pur-chased LNG from North Sumatra to the UnitedStates (8,300” nautical miles each way), Pac In-donesia has entered into contracts for the hire

Table 27.—Estimated Capital Requirements forEl Paso Atlantic—Six Vessels

(thousands of 1976 dollars)

of nine vessels. Three of the vessels have al-ready been constructed in foreign shipyards,and plans call for the remaining six to be built inU.S. yards. All of the ships would be available toPac Indonesia under time charter agreements,which provide for monthly billing beginning onspecified dates. The FPC’S administrative lawjudge described these arrangements as follows:


Table 28.—Estimated Annual OperatingExpenses for El Paso Atlantic—Six Vessels

Thousands ofOperating expenses for six vessels 1976 dollars

Crew (3 foreign, 3 U.S.). . . . . . . . . . . . . . . . . . . . . .$ 8,172Maintenance and repair. . . . . . . . . . . . . . . . . . . . . 5,856Stores and supplies @ $103,000/ship . . . . . . . . . 618Bunker “C” fuel @ $1,371,000/ship . . . . . . . . . . . 8,226Nitrogen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354Annual insurance premiums. . . . . . . . . . . . . . . . . 8,484Post charges. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2,484Shoreside expenses . . . . . . . . . . . . . . . . . . . . . . . 3,527Manning agent (3 foreign ships only). . . . . . . . . . 42Miscellaneous expenses @ $29,000/ship . . . . . . 174

Estimated total annual operating expenses ... .$37,937

SOURCE. El Paso Atlantlc Company, Economics of Shipping A/ger/an LNG toTexas Gulf Coast, Oct 11, 1976

accident or damage to the vessel, breakdown ofthe vessel’s machinery, deficiency of men or

stores). These risks are thus borne by the ship-owners, not Pac Indonesia. 16The capacity of each vessel will be about

125,000 ms, the industry’s current standard.Each carrier will be scheduled to operate 345days out of the year (as opposed to 332 to 333 inthe Algeria II project) leaving the balance of theyear for shipyard repairs and miscellaneous de-lays. At an average speed of 18.5 knots, eachship will require 18.7 days for the 8)300 nauti-cal-mile voyage from Indonesia to the UnitedStates and will be able to complete 8.5 roundtrips per year. The energy delivered to theUnited States will be approximately 92 percentof the quantity loaded at the A run terminal, re-flecting fuel use of boil-off vapors during thevovage..

To illustrate how the shipping distance affectscosts, the capital and operating costs per vesselon Pac Indonesia’s project are assumed to equalthe corresponding costs for Algeria II. Undersuch an assumption, the shipping costs per mil-lion Btu of LNG delivered daily in the Pac Indo-nesia project would exceed by 42 percent theequivalent costs in the Algeria 11 case, as a resultof the greater distance involved in the Inck)-nesian project.

The capital costs of the three foreign shipsconstructed in French shipyards* and char-

tered by Pac Indonesia are not publicly dis-closed, but can be reasonably estimated at ap-proximately $100 million per vessel. Pac Indo-nesia’s transportation contracts with U.S. ship-pers, concluded in late 1975, provide for char-ter rates based, in part, on the estimated capitalcosts of $140 million per American-mack tinselplus escalations. The FPC judge used $155 mil-lion estimated average capital cost per U.S. ves-sel in establishing the shipping component ofPac Indonesia’s initial certificate rate. Thisfigure represents the judge’s estimate (inmid-1977) of the average cost for the six U.S.tankers assuming a specified delivery schedulebetween January 1980 and May 1981. The actu-al inflation in LNG tanker construction costs inthe United States turned out to be higher, judg-ing by the current (mid-1979) total price esti-mate of about $195 million (after subsidy) per125,000 m3 vessel to be delivered in 1982-83.

Receiving country terminal andregasification facilities

A terminal for the receipt of LNG consists ofthree major segments -unloading, storage, andvaporization. The principal components of theunloading portion of the terminal are berthingfacilities, unloading arms and lines, returnvapor lines and blowers, and prov’isions for han-dling excess vaporization due to boil off. Thestorage facilities at the receiving terminal aresimilar in type to those at the liquefaction plant.“The regasification (vaporizing) equipment con-sists of liquid pumps and vaporizers. Regasifica-tion facilities represent much less sophisticatedtechnology than do the liquefaction plants in theproducing country. In terms of the total costs in-volved, the importing country’s facilities usuallyaccount for the smallest portion of a three-partLNG project. The design and construction of thereceiving terminal facilities require 2 to 3 years.

Pac Indonesia proposes the construction of anLNG terminal on the southern California coastapproximately 3.5 miles east of Point [concep-tion. * The plant will have an ultimate baseloadcapability of 1,300 M Mcf/d, with peak \’aporiza-.

86 c The Future of Liquefied Natural Gas Imports

Photo credit’ Courtesy of Colomb/a Gas System, Inc.., Consolidated Natural Gas Co, and American Petroleum Institute

LNG receiving terminal at Cove Point, Md. At the terminal LNG will be converted back into ordinary natural gas foruse by customers of the Columbia Gas System, Inc., and the Consolidated Natural Gas Company

Ch. 4—Project Structure, Cost, and Financing . 87

tion capacity of an additional 300 MMcf/d. TheIndonesian volume to be received at the termi-nal is estimated to be about 500 MMcf/d. The to-tal baseload capacity will be shared by Pac Indo-nesia and Pacific Alaska LNG Associates (the lat-ter proposes to import LNG from the Cook Inletarea of Alaska).

The marine facilities will consist of one berthlocated about 4,600 ft offshore. LNG from theship will be unloaded into the land-based LNGstorage tanks by onboard ship pumps. Three550,000 barrel double-wall, insulated storagetanks are planned for the terminal.

Thirteen seawater-heated LNG vaporizers willbe installed to accommodate the total baseIoad,and peaking capacity of 300 Mcf/d will be pro-vided by additional gas-fired LNG vaporizers.The vaporization plant is designed to deliver na-tural gas continuously 365 days per year. Thegas will then go through a trim heater, odoriz-ers, and metering station, before entering thegas transmission system.

A 112-mile, 34-inch pipeline looped withanother 45-mile, 34-inch pipeline will extendfrom the metering station at the terminal site toa point of interconnection with PG&E’s existingpipeline near Gosford, Calif., with an interven-ing interconnection with Southern CaliforniaGas Company’s present facilities at North ColesLevee. The present pipeline design requires nocompressor stations. 17

The total estimated capital cost of the PointConception terminal amounts to $632 million (inmid-1977 dollars), and the annual operatingcosts to $20 million—see tables 29 and 30 re-spectively. The estimated capital cost of the newpipeline requires another $117 million (see table31). On a strictly volumetric basis, the Pac Indo-nesia share will be over 58 percent, or $368 mil-lion for the terminal facilities and $68 millionfor the pipeline. Pac Indonesia’s costs would behigher if the facilities were built for the use ofthis project alone. For instance, all storage tankswould still be needed, due to the industry’spractice of requiring that storage space be suffi-

88 ● The Future of Liquefied Natural Gas /reports

cient to accommodate at least two LNG ship-loads. Pipeline costs also exhibit economies ofscale.

The Algeria 11 project involves the construc-tion of the La Salle terminal in Matagorda Bay,designed for a maximum sendout rate of 1.64Bcf/d. Thus, unlike Pac Indonesia’s terminal, LaSalle would be serving only the Algeria 11 proj-ect. 18 The marine terminal consists of independ-ent berths to accommodate two LNG carriers si-multaneously. More storage will be availablethan in Pac Indonesia’s terminal; three 629)000barrel tanks. The estimated cost of the La Salleterminal is $456 million (4th quarter, 1976),which as table 32 indicates, amounts to a highercost per million Btu delivered than in Pac Indo-nesia’s project. This discrepancy is due primar-

ily to the volumetric cost allocation for the PacIndonesia project, and only secondarily to thephysical differences between the two terminals.

Algeria II requires more extensive pipeline fa-cilities on the receiving end than does Pac Indo-nesia. El Paso Natural proposes to build a pipe-line capable of accepting 115 percent of the av-erage daily output of La Salle terminal, or 1,065MMcf. The first 31 miles of the new pipeline (36-inch diameter) will transport the gas from theLa Salle terminal to United LNG’s present facili-ties near Victoria, Tex., where 35 percent of thetotal quantity will be sold. The remaining 65percent will be transported via a 432-mile (30-inch diameter) pipeline to El Paso Natural’s sys-tem at Waya, Tex. Together with the requiredfive compressor stations, the new pipeline facili-ties are estimated to cost $263 million. As shownin table 32, the pipeline cost per million Btu perday is 37 percent of the total capital investmentin Algeria II import facilities, whereas similarcosts for the Pac Indonesia project are only 16percent.

Table 32.–Capital Costs for Import Facilities per Million Btu of Daily Delivered Quantity of LNG

LNG financingBecause much of the cost of an LNG project is

incurred at the beginning of the project, and be-cause an LNG project has a long economic life-time, financing terms strongly influence theunit cost (cost-of-service) of moving the gas fromthe field to the market. This section examinessome of the major financing options open toLNG project sponsors and then, incorporatingthis information, derives an idealized cost-of-service.

Overview

The fundamental determinants of financibil-ity for any capital project are risk and return.The important characteristics of an LNG projectaffecting perceptions regarding risk and returnare:

● The total capital costs of an LNG project arelarge, and the return is not certain.


●

●

Ownership of LNG projects is often spreadamong parties in different countries.Integrated LNG projects are comprised ofseveral stages (e.g., liquefaction, shipping),each one possessing an individual identity.

To see how the scale of capital requirementsfor an LNG project impacts financibility, it isuseful to put the capital requirements in per-spective. The estimated costs of the proposedAlgeria 11 project total over $5 billion from well-head to final consumer. By comparison, totalU.S. net private fixed investment in 1978 wasonly $128.7 billion, * and this for the largesteconomy in the world.

Photo credit. El Paso Co

Natural gas transmission lines may be seen above--ground in remote areas, but most of the Nation’s

pipeline system is underground

Because capital requirements are so large,project sponsors may have to look to severalcapital markets for funding, simply because thetotal exposure would be too large for one mar-ket to absorb. By using several capital marketsand many lenders, project sponsors can diffusethe large financial risk of the project andthereby reduce borrowing costs. However, theuse of several capital markets (or, for that mat-ter, large financing in one capital market) mayentail substantial transactions costs, offsettingthe gains achieved through this strategy and, infact, ultimately limiting the degree of diversifi-cation that is economically feasible.

Transaction costs incurred through diversifi-cation may take several forms. The requirementfor documentation alone can be significant. Inthe U.S. institutional market, for example, eachof several separate bond issues underwrittenand offered publicly could require a separateprospectus and indenture, demanding signifi-cant outlays for legal, accounting, and possiblytechnical services. Transactions costs may alsotake the form of decreased flexibility. Restric-tive covenants required in one capital marketon, say, an issue of unsecured bonds may limitthe project sponsors’ freedom in obtaining fi-nancing in other markets.

In addition to the high transaction costs of re-liance on many sources for financing, the size ofthe project ultimately limits the capacity of capi-

90 ● The Future 0f Liquefied Natural Gas Imports

tal markets to absorb the risk. Whereas a mod-estly sized capital investment can be dividedamong many investors so as to represent only asmall portion of any single portfolio, an LNGproject is sufficiently large that enough lendersmay not be available to distribute the risk ade-quately.

A second major factor influencing financi-bility is the international character of LNGprojects, since financiers look to the contractsamong the parties for security. Since the con-tract signatories are typically domiciled in dif-ferent countries and, perhaps more fundamen-tally, since their physical facilities are located indifferent countries, no one legal jurisdiction canenforce the claims of one party against another.

A third factor is the multistage nature of theLNG project, and separate stages of the LNGproject may have access to different capitalmarkets for several reasons. First, for facilitiesto be owned, for example, by the producingcountry, officially supported credits may beavailable from countries desiring to promote ex-ports from their own construction and capitalgoods industries. A second reason is that poten-tial lenders may have different attitudes towardrisk depending on the stage. An LNG import ter-minal, for example, can be used efficiently forone purpose only: the receipt, storage, and re-gasification of liquid gases at the location wherethe terminal is built. If the project fails becauseof, say, market conditions in the importingcountry, the terminal just sits there generatingcapital charges. The LNG carriers, on the otherhand, if prohibited from offloading at the inop-erative terminal can still be used in an L N Gtrade somewhere else. Thus, the potential in-vestors might perceive less risk attendant onLNG carriers than on an import terminal.

The following sections examine, in light ofthese general considerations, some of the fi-nancing options open to the sponsors of LNGprojects, with a particular view toward their ef-fects on project cost. The discussion is orga-nized by production stage: first, exporting coun-try facilities, then ships, and finally, U.S. importterminals. For each stage, the discussion of fi-nancing focuses on the debt requirements, and

the section on the financing of exporting coun-try facilities includes overall project equity.

Exporting country facilities

Total capital requirements for exporting facil-ities may vary considerably with differences ingasfield characteristics, distance of field fromthe plantsite, cost of local labor, and other varia-bles, but for a project of 1 Bcf/d the total cost ofall exporting country facilities is likely to be wellover $1 billion, and may exceed $2 billion.

The total cost of the facilities may be financedwith the credit backing of the exporting countryitself, as in the case of the Algeria 11 project, byoutside participants, such as multinational oilcompanies, or a combination of both. While“project” financing, for which the security is thevalue of the specific facilities or contractual obli-gations associated with the project itself, may bepossible in concept, financing is not likely to beobtained in this way without independent creditsupport.

Major sources of capital for producing coun-try facilities include the eurocurrency market,private and public equity, and in the case of ex-porting country ownership, officially supportedexport credits. Each of these sources are dis-cussed below.

OFFICIALLY SUPPORTED EXPORT CREDITS

Several Organization for Economic Coopera-tion and Development (OECD) countries haveofficially supported export credit programs,which supply direct loans and credit guaranteesto promote their industries. This tied financingoffers some important advantages to the LNGexporting country. While some export creditprograms, such as those of the United Statesand Germany have tended to be on basicallycommercial terms, other countries, such asFrance, Japan, and the United Kingdom offerpreferential–if not concessionary–credit sup-ports. The lower cost of the financing availablefrom some countries improves the economic vi-ability of the project from the point of view ofthe producing country and also lowers the cost-of-service. A second advantage of officially sup-ported export credits is that other potential


lenders to the project may feel more secure intheir investments with the participation of of-ficial government agencies, and in any event,will perceive that the lower cost of fundsavailable through export credit financing pro-vides additional capacity to service private debt.

France, through its foreign trade bank BFCE(Banque Francaise du Commerce Exterieur),provides financing for long-term maturitiesthrough either direct credits at subsidized ratesor through discount and refinancing arrange-ments. The rate on the BFCE direct credit, ordiscount on the subsidized portion of the bankloan, is set by BFCE so that the blended rate, *exclusive of fees and premiums, is at the mini-mum allowed under the OECD arrangement. * *In addition, Coface (Compagnis Francaise d’As-surance pour le Commerce Exterieur) guaran-tees the total amount of the credit (BFCE plusprivate portion) for a rate premium of approx-imately 0.85 percent.

In some cases, however, the French “mix”credits by tying aid and loans together in onepackage. Such tied-aid credits may include loansat rates as low as 3.5 percent with repaymentterms up to 20 years. The average cost of such apackage is therefore considerably less than itwould be in a strict export finance deal.

Japan, through the Ministry of InternationalTrade and Industry and its Export-Import Bank,also provides long-term finance packages on

preferential terms. As France’s BFCE, Japan’sExport-Import Bank will extend direct credits,up to 60 percent of the total export financing, sothat the blended rate is at or near the minimumallowed by the OECD arrangement. As withCoface, their Export-Import Bank provides in-surance for its own and the private loan por-tions of the total credit for a premium.

Japan has officially denied that it offers tied-aid credits to an extent that would derogate theOECD arrangement. However, some claim thattied-aid credits on essentially concessionaryterms are widely available for export financingfrom Japan.

The United Kingdom through ECGD (ExportCredits Guarantee Department) sets rates forcommercial bank loans to buyers and pays a di-rect interest subsidy to banks to make up forthe actual cost of funds. To relieve the overallcredit burden this creates, ECGD also provideslimited refinancing for sterling denominatedloans used for export financing. ECGD guaran-tees 100 percent of the bank loans, and the ratesset by ECGD are, as in the case of France andJapan, at or near the minimum allowed underthe OECD arrangement. British tied-aid credit fi-nancing is also available.

The United States and Germany are more con-servative in their approach to export financing.Historically neither country has typically of-fered tied-aid financing of exports. In addition,both countries operate their respective exportprograms without recourse to subsidy–in thecase of Germany only a modicum of direct cred-its are even provided at long-term and preferen-tial rates, the bulk of long-term credit supporttaking the form of insurance.

The United States through the Export-ImportBank (Eximbank) and related organizations suchas the Private Export Funding Corporation, pro-vides support for long-term export credits. His-torically, Eximbank has provided direct credit atrates linked to the agency’s cost of funds andhas guaranteed the private bank portions of thetotal credit, which are usually extended at float-ing rates. Eximbank typically charges a guaran-tee fee on the private bank portion.


Recently, however, U.S. Eximbank policy hasbegun to favor improving competitiveness withforeign export agencies. This policy shift is re-flected in the tendency toward increased directcoverage at reduced rates. In so-called excep-tional cases, Eximbank may offer a direct creditfor the total amount of the export credit (85 per-cent of the contract cost of the goods and serv-ices) at rates below the agency’s marginal cost offunds.

An example of the use of export credit is theU.S. Eximbank’s $240 million credit to Algeria tohelp finance $320 million U.S. goods and serv-ices component of the Arzew 11 liquefactionplant. The credit was extended to Sonatrach at8.5 percent. Repayment is in 20 semiannual pay-ments beginning 6 months after the last of sixliquefaction trains is completed. The remainderof the U.S. goods and services component is tobe financed by a Sonatrach payment of $48 mil-lion (15 percent) and private-source loans of $32million. Payments on the total of the Eximbankcredit and the private source loans will be ar-ranged in such a way that the private-sourceloans are paid off first.

THE EUROCURRENCY MARKET

Another important source for financing ofLNG exporting country facilities (whetherowned by the country in question or by outsideparties) is the eurocurrency market, in whichloans are negotiated in currencies not native tothe country in which the bank offering the loanis located (eurocurrency bank credits) andbonds are issued outside the country of the bor-rower (international bonds). The size of thismarket is substantial. In 1978 alone, over $70billion in eurocurrency bank credits were nego-tiated, and new international bond issues to-taled $35 billion. In 1978, Algeria, a major LNGproducing nation, borrowed over $3 billion onthe eurocurrency market, while Indonesia,another important LNG center, borrowed over$1 billion.

An important feature of the eurocurrencybank credit market is that its funding tends tobe for periods of no more than 5 to 10 years.Also, loans on this market typically have floatinginterest rates. So, for example, a loan on thismarket with a repayment term of 8 years might

carry an interest rate of 1.75 percent overLIBOR (the London interbank offering rate, ameasure of the bank’s cost of funds), reflectingmaturities of 6 months. At the end of each 6-month period, the loan is effectively renewedfor the amount of principal still outstanding atan interest rate corresponding to the then- cur-rent LIBOR.

One of the main advantages of the eurocur-rency bank credit market is that with sufficientcredit backing, such as the guarantee of theCentral Bank or Development Bank of the po-tential LNG exporting country, considerablefunds are available on this market. A second ad-vantage is that credit obtained on this market,and private-source capital in general, tends tohave fewer strings attached than, for example,the tied loans available through officially sup-ported export-financing agencies.

Two important disadvantages of this marketare the shortness of the repayment periods andthe variability of the interest rates. The econom-ics of large projects with long lifetimes some-times are not certain until late in the project,and if the loan must be amortized over too shorta period at the beginning, debt service may ex-ceed the available cash flow after deduction ofother expenses. In such an instance, borrowingfrom equity is required, and to the extent thatthis is expensive or simply not feasible, otherfinancing arrangements are necessary,

The variability of interest rates on eurocur-rency bank credit also adds a dimension of un-certainty to the management of cash flow and tothe overall economics of the project. owners oflong-term projects may be willing to pay a con-siderable premium to remove this element ofuncertainty.

An example of eurocurrency bank credit isthe $250 million 7-year loan raised by Sona-trach, guaranteed by the Banque Nationalede’Algerie, and jointly led by Citicorp, Bank ofAmerica, Apicorp, Bankers Trust, Bank of Mon-treal, and Continental Illinois. This loan carriesan interest rate of 1-3/8 percent over LIBOR andwill help to finance the Arzew II liquefactionplant facilities.


Fixed interest rate financing is also possible onthe euromarket through the issue of bonds ornotes. These international bonds, which arecomparable in terms of maturity with eurocur-rency bank credits (5 to 10 years), can provideadded cash flow predictability at a minimumcost.

Examples of eurobonds are two recent Sona-trach issues, one 12 million dinars (DA) guaran-teed by the Banque Exterieure d’Algerie, matur-ing in 10 years, bearing a yield of 8.5 percent;and one DA 8 million 5-year maturity bearing8.5 percent.

PUBLIC AND PRIVATE EQUITY

A third important element of financing forLNG exporting country facilities is public andprivate equity. Equity capital can be generatedby reinvestment of earnings, such as profitsfrom a country’s other hydrocarbon ventures orthe net revenues from unrelated operations of amultinational oil company; or alternatively,through the issue of ownership shares, as ex-emplified by the Islamic Development Bank’sequity participation in Jordan’s petroleum re-finery project. Equity can be public, resultingfrom taxation or earnings on public enterprises;or private, supplied by ownership of shares byprivate entities.

Generally, lenders require some equity as abuffer in the event of difficulties, but it candilute ownership and is typically more expen-sive than debt. The U.S. Eximbank, for exampIerequires a 15-percent cash payment by thebuyer of U.S. export goods. This 15-percent maybe funded by equity or debt or both, however,and this 15-percent should therefore not beviewed as necessarily bearing “true” equitycosts.

CONCLUS1ONWhether the project owners are to be the ex-

porting country itself, an outside entity, or acombination, many financing options are availa-ble as described above. Nevertheless, certainconstraints should be recognized. The availabil-ity of officially supported export credit financ-ing at preferential or concessionary rates maydepend on who owns the producing country fa-cilities. In addition, while project financing is

possible in principle, it is rare in practice, andconsequently, the total cost of the financingmay exceed the prima facie cost of the borrow-ings because of the impact on the debt capacityof the guarantor.

Shipping

Many of the private capital markets for the fi-nancing of exporting country facilities are openalso for LNG ships. In addition, as in the case ofthe exporting country facilities, officially sup-ported financing is available for LNG shippingand may be provided by export credit agenciessuch as Germany’s Hermes (Hermes Kreditver-sicherung) or by other government agenciessuch as the U.S. Maritime Administration(MarAd).

MarAd offers a loan guarantee program ap-plicable to LNG ships if they are built in U.S.shipyards, registered under the U.S. flag,owned by U.S. entities, and crewed by U.S. citi-zens. Under these conditions, as much as 87.5percent of the total cost of a ship can be fi-nanced by an issue of U.S. Government guaran-teed serial or sinking fund bonds, either withmaturities up to 25 years. The cost to the bor-rower is the yield on the bonds plus the MarAdguarantee fee (0.5 to 1.0 percent of the out-standing balance), amounting to a total in therange of a Baa industrial.

The MarAd program is not subsidized, thoughdefault claims are paid from a pool funded byMarAd’s overall operations. Consequently, thepublic does not contribute directly to defrayingthe cost of funds to borrowers using MarAdcredit guarantees. However, the MarAd guaran-tee is valuable to potential borrowers in thesense that MarAd may be better able to spreadthe risk in ship financing than private capitalmarkets. In addition, the ability to issue U.S.Treasury-backed bonds affords shipowners ac-cess to other capital markets that, because of in-stitutional or legal barriers, would otherwise beclosed. Another advantage of the MarAd pro-gram is long repayment terms of up to 25 years,although the maximum term might be unusualin the case of a LNG ship.

If the owners of the ships are to be foreign tothe country where the ships are built, officially


supported export credits may be available. Im-portant LNG ship-exporting countries in theOECD include France, the United States, Italy,and Norway. Financing terms are determinedby the individual countries in accordance withthe OECD Arrangement on Ships, which super-sedes the Arrangement on Guidelines for Offi-cially Supported Export Credits in the case ofshipping and sets a minimum interest rate of 7.5percent, a maximum repayment period of 8years, and a maximum coverage of 80 percentof ship cost.

While the interest rate allowed on officiallysupported export credits for ships is preferen-tial, the shortness of the repayment period con-stitutes a disadvantage of this type of financing.The heavy debt requirements during the earlyyears of the project can, depending on the pric-ing or tariff provisions governing cash flow,effectively postpone repayment of expensiveequity capital.

U.S. facilities

The cost of U.S. facilities, including themarine terminal, LNG storage tanks, vaporiza-tion units, and transmission lines to deliver thegas, can vary widely depending on location anddesign. The proposed La Salle terminal facilitiesand delivery lines to E] Paso United’s systemwere estimated to cost approximately $700 mil-lion, while the proposed Tapco project, spon-sored by Tenneco to bring LNG from Algeriainto New Brunswick, Canada and then to theUnited States through a longer pipeline, wouldhave cost nearly $1.5 billion.

In the past, construction of U.S. facilities hasbeen accomplished through the credit of finan-cially strong corporations. Examples includeSouthern Natural Gas Company’s Savannah, Ga.,terminal and the planned Trunkline terminal atLake Charles, La. Each of these facilities isowned and operated by a subsidiary of a majorU.S. interstate pipeline company, and debt is-sues to finance LNG terminals are carried in

both cases on the balance sheets of the parents,which provide substantial credit support.

Southern’s Savannah terminal illustratesanother feature of U.S. financing, the tax-exempt bond market. Section 103(B)(4)(D) of theU.S. Internal Revenue Code, which relates todocks, wharfs, and storage facilities, may allowthe issue of debt securities that are exempt fromFederal, State, and local income taxes. For thistype of issue, the market yield cost to the bor-rower is less. Southern’s Savannah terminal, forexample, was financed partly through the issueof tax-exempt revenue bonds by the SavannahPort Authority. These bonds were marketed ata price of 99.75 percent of par and with a cou-pon rate between 5.7 and 6.75 percent, depend-ing on the maturity (serial and sinking fundbonds were in combination to permit full amor-tization by equal payments over the lifetime ofthe financing). At the time of the prospectus,Aaa bonds were yielding around 9.5 percent,considerably more than the Savannah Port Au-thority bonds.

Conclusion

The financing options open to LNG importsponsors are strongly influenced by the magni-tude of total capital requirements, as well as theinternational character and multistage natureof the projects. Generally, private capital mar-kets are open to sponsors with strong creditsupport, but financing costs may be high be-cause of sheer scale.

In addition, low-cost subsidized financing isavailable for particular stages of the project,depending on such factors as ownership, loca-tion, the country supplying construction andmaterials, and taxes. Also financing costs maybe high due to regulatory incentives, for exam-ple, to finance U.S. facilities independently. Thenext section examines the effects of alternativefinancing arrangements on the unit cost (cost-of-service) of moving gas from a remote sourceto U.S. markets.


Cost-of-serviceThe cost-of-service calculated by project stage

for a hypothetical world-scale LNG project (ap-proximately 1 Bcf/d), idealized in economic andfinancial structure, is described below. This costwas defined as the sum, for a particular year, ofoperating costs, capital costs (debt and equity),and taxes divided by the Btu’s delivered into theU.S. pipeline system. The procedure alloweddirect comparison by project-stage in terms ofunits ultimately delivered, and at the same time,allowed determination of the wellhead value ofthe gas given the price in the U.S. market, sincefuel and loss is included in the cost-of-service.

The capital charge (service on debt and eq-uity) was assumed to be level in constant dollarsover the lifetime of the project and was calcu-lated so that the net present value of the share-holders’ cash flow would equal zero at the as-sumed discount rate.

Also, the cost-of-service estimate here differsfrom what would be typical in a regulatory pro-ceeding in that the cost of gas expended as fueland loss was determined at the ultimate netbackwellhead value, while a “regulatory” calculationusually assumes the value of fuel and loss to re-flect a fixed price, for example, the contract saleprice f.o.b. the point of export. An advantage ofvaluing the fuel and loss at the ultimate well-head netback, is that it allows direct comparisonof the cost-of-service of LNG projects with thecost-of-service of alternative technologies tomove gas from the wellhead to the market. Apeculiarity of this approach is that since capitaland nonfuel operating costs affect the ultimatewellhead netback, the value of fuel and loss de-pends, in part, on these other components.

Base case

Parameter values for the base case cost-of-service calculation were derived largely fromthe testimony in recent LNG proceedings. This

calculation was performed for a project startingin 1985, and costs were escalated to reflect thedifference in timing between the idealized basecase and recent proposals. The assumed 3,736nautical miles approximates the shipping dis-tance from Algeria to the United States. Forcomplete enumeration of all parameters, seevolume II (working papers).

For U.S. facilities, the financing was assumedto reflect current market rates and terms forlarge projects in U.S. capital markets on the bal-ance sheets of gas utility companies. For ship-ping, MarAd financing was assumed with an in-terest rate in the vicinity of Baa industrial bondsand a repayment term of 15 years. Exportingcountry facilities were assumed to hav’e accessto export credit financing programs for the bulkof their financing, and rates and terms repre-sented here are those for the recent U.S. Exim-bank credit to Algeria to help finance the ArzewII liquefaction plant. The assumed rate of returnon equity is 17 percent.

Tables 33 and 34 show the results of a repre-sentative base case cost-of-service calculation,expressed in constant 1978 dollars. As can beseen, the bulk of the costs are in shipping andliquefaction. Also fuel and loss represent a sig-nificant portion of the total and, for liquefac-tion, actually exceed the capital and operatingcost component. Overall, taxes constitute lessthan 10 percent of the fifth year cost-of-servicefor this idealized project. However, much of thedebt service is loaded toward the front end ofthe project and since interest payments are de-ductible, taxes as a fraction of the total cost-of-service increase over the lifetime of the project,as shown in figures 17 and 18.

Sensitivity

The sens i t iv i ty of the cost-of-service tochanges in the debt ratio on U.S. facilities andthe repayment period for ship financing, and to


Table 33.—Cost of Service of an LNG Project Beginning in 1985 in the Fifth Year of Operation”(1978 dollars/million Btu delivered into U.S. pipeline system)

Capital and operating . . . . . . . . $0.226 $0.656 $0.564 $0.273 $1.719Fuel and loss . . . . . . . . . . . . . . . 0.084 0.659 0.091 0.101 0.935Income taxes . . . . . . . . . . . . . . . 0.053 0.106 0.038 0.017 0.214

Total cost-of-service. . . . . . . $0.363 $1.421 $0.693 $0.391 $2.868

statute miles

SOURCE Jensen Associates, Inc.

Figure 17.— LNG Cost of Service by Project Section

3.00

t

2.00

1.00

0 [1;84 85 86 87 88 89 90 91 92 93 94 95

Year

SOURCE Jensen Associates, Inc

setting the interest rate on all debt at 12 percentis relatively insignificant. These changes pro-duced a net cost increase of $0.05 /MMBtu. How-ever, the cost-of-service is more sensitive tochanges in the return on equity in all portions ofthe project and to changes in the U.S. debt inter-

4.00

3.00

2.00

[

1984 85 86 87 88 89 90 91 92 93 94 95Year

SOURCE. Jensen Associates, Inc.

est rate, Tables 35 and 36 show the effect of in-creming and reducing the return on equity andthe interest on U.S. debt by 2 percent. The ef-fects are symmetrical, changing the totalcharges, excluciing fuel and loss, by about$0.21 /MMBtu (1978 dollars) or approximately 10percent. However, when the fuel effects, whichoffset the gross changes in capital and operatingcost portions of the project, are included, thenet effect is a change of about $0.16/MMBtu(1978 dollars) or approximately 6 percent in thetotal cost-of-service in the fifth year of operationof an LNG project. Cost-of-service calculationsfor shipping were based on an average distanceof 7,472 nautical miles. Tables 37 and 38 showthat when the distance is reduced by one-half ordoubled, the effect on the cost of shipping is inthe same proportion. However, the netback val-


Table 35.—impact of Fifth Year Cost of Service of Reducing Return onEquity to 15 Percent and Interest on U.S. Debt to 10 Percent”

(1978 dollarsl million Btu ultimately delivered)

Field and pipelineU.S. facilities Shipping Liquefaction plant facilities Total

Reduced return capital costs,operating costs, and incometaxes. . . . . . . . . . . . . . . . . . . . $0.246 $0.546 $0.691 $0.244 $1.727

Base case capital costs,operating costs, and incometaxes. . . . . . . . . . . . . . . . . . . . 0.290 0.602 0.763 0.279 1.934

Gross change in cost ofservice . . . . . . . . . . . . . . . . . . (0.044) (0.056) (0.072) (0.035) (0.207)

Reduced return fuel and loss. . 0.106 0.095 0.692 0.088 0.981Base case fuel and loss . . . . . . 0.101 0.091 0.659 0.084 0.935

Change in fuel and losscomponent. . . . . . . . . . . . . . . 0.005 0.004 0.033 0.004 0.046

Net change in cost of service. . (0.039) (0.052) (0.039) (0.031) (0.161)

Table 36.—impact of Fifth Year Cost of Service of Increasing Return onEquity to 19 Percent and Interest on U.S. Debt to 14 Percent”

(1978 dollarsl million Btu ultimately delivered)


Raised return case capitalcharges, operating costs,and income taxes . . . . . . . . . $0.337 $0.662 $0.833 $0.314 $2.146

Base case capital charges,operating costs, and incometaxes. . . . . . . . . . . . . . . . . . . . 0.290 0.602 0.763 0.279 1.934

Gross change in cost ofservice . . . . . . . . . . . . . . . . . . 0.047 0.060 0.070 0.035 0.212

Raised return case fuel andloss . . . . . . . . . . . . . . . . . . . . . 0.096 0.086 0.626 0.080 0.888

Base case fuel and loss . . . . . . 0.101 0.091 0.659 0.084 0.935

Change in fuel and losscomponent. . . . . . . . . . . . . . . (0.005) (0.005) (0.033) (0.004) (0.047)

Net change in cost of service. . 0.042 0.055 0.037 0.031 0.165

aFrom 17 and 12 percent respectively.

SOURCE. Jensen Associates, Inc

ue of the gas and thus the cost of fuel are re-duced, which tends to offset the changes in cap-ital and operating costs. Another effect of short-ening shipping distance is a reduction in theamount of LNG boiled-off and used as transpor-tation fuel. Consequently, less LNG needs to beloaded at the liquefaction plant to maintain thesame deliveries to ultimate destinations. For thisreason, liquefaction plants and field and pipe-line facilities can be reduced somewhat in scale,as shown in table 37. on the other hand, in-

creasing shipping distances will produce the op-posite effects.

Ironically, the total fuel and loss portion of thecost-of-service for the short voyage case is high-er than for the base case and similarly, total fueland loss for the long voyage case is lower thanfor the base case. This result is contrary to whatone would expect but comes about because ofthe convention adopted in this analysis, that thecost of fuel and loss is calculated at the netback

S9-406 O - 80 - 8


Table 37.—impact on Fifth Year Cost of Service of Reducingthe Round Trip Voyage Distance to 3,274 Nautical Miles’

(1978 dollars/million Btu ultimately delivered)


Short-voyage capital charges,operating costs, and incometaxes. . . . . . . . . . . . . . . . . . . . $0.290 $0.300 $0.754 $0.276 $1.620


Gross change in cost ofservice . . . . . . . . . . . . . . . . . . . — (0.302) (0.009) (0.003) (0.314)

Short-voyage fuel and loss. . . . 0.110 0.053 0.709 0.090 0.962Base case fuel and loss . . . . . . 0.101 0.091 0.659 0.084 0.935Change in fuel and loss

component. . . . . . . . . . . . . . . 0.009 (0.038) 0.050 0.006 0.027Net change in cost of service. . 0.009 (0.340) 0.041 0.003 (0.287)

aBase case distance IS 7,472 nautical miles.

SOURCE: Jensen Associates, Inc

Table 38.—impact on Fifth Year Cost of Service of Increasingthe Round Trip Voyage Distance to 16,694 Nautical Miles’

(1978 dollars/million Btu ultimately delivered)


Long-voyage capital charges,operating costs, and incometaxes. . . . . . . . . . . . . . . . . . . . $0.290 $1.294 $0.784 $0.287 $2.655


Gross change in cost ofservice . . . . . . . . . . . . . . . . . . —— 0.692 0.021 0.008 0.721

Long-voyage fuel and loss . . . . 0.082 0.149 0.546 0.070 0.847Base case fuel and loss . . . . . . 0.101 0.091 0.659 0.084 0.935

Change in fuel and losscomponent. . . . . . . . . . . . . . . (0.019) 0.058 (0.1 13) (0.014) (0.088)

Net change in cost of service. . (0.019) 0.750 (0.092) (0.006) 0.633

value at the wellhead. Since most of the fuel is value at the wellhead and thus the cost of fuelused in the liquefaction plant, the short voyage and losses. When the shipping distance is re-raises the netback value of the fuel and loss duced by half the total cost-of-service for thewhich in turn increases the fuel and loss prices project is reduced by 10 percent, and when theat the liquefaction facility. The opposite is true distance is doubled, the cost-of-service is in-when LNG is shipped over long distances. High- creased by 22 percent.er capital costs of boil-off reduce the netback

Ch. 4—Project Structure, Cost, and financing 9 9

Financial riskBecause of the size and complexity of LNG in-

vestments as well as the unpredictability of fu-ture events during a project’s life, the natureand distribution of financial risk and uncertain-ty are important to consider. From a govern-ment policy standpoint, public exposure de-serves particular attention.

Individual investments in single liquefactionfacilities, import terminals, or ships separatefrom integrated projects are not yet possible,except perhaps in Japanese trades. In the fu-ture, if more trade develops and facilities be-come more widespread, such investments willoccur. But now only a complete project withlong-term contractual commitments and simul-taneous construction of liquefaction facilities,ships, and the receiving/regasification terminalis economically feasible.

The money invested in LNG projects is at riskfrom a variety of perils, including technical feas-ibility, project failure, project interruption ordelay, cost overruns, and market uncertainties.These problems are described below and illus-trated in a discussion of the proposed Pac-Indo-nesia project.

Technical feasibility

The financiers of an LNG project are first in-terested in its technical feasibility, which theyassess in historical terms. In January 1959, thefirst shipload of LNG left Lake Charles, La., onthe Methane Pioneer and arrived at CanveyIsland, England—near London. This shipmentwas part of a project sponsored by the BritishGas Council, Continental Oil Company, andUnion Stockyards to test equipment and to ex-amine the feasibility of international LNG trade.In 1964, the first baseload, long-term commer-cial project started with shipments from Arzew,Algeria, to Canvey Island and Le Havre, France.Today 12 LNG projects are in operation, and thetechnical and economic feasibility of the tech-nology has been demonstrated. Several ship-yards and engineering/construction firms haveshown their ability to build reliable LNG shipsand facilities. Technical and construction risks

are perceived as no greater than those for otherlarge sophisticated international engineeringand construction projects.

Project failure

Of greatest concern to the owners and finan-ciers of an LNG project is the possibility for proj-ect failure after significant amounts of moneyhave been spent. For example, the EascogasProject, which was to bring LNG from Skikda,Algeria, to the east coast of the United States,was originally expected to begin some yearsago. A receiving terminal at Staten Island wascompleted at a cost of over $100 million (1974dollars), and two ships intended for the projectwere built at the General Dynamics shipyard atQuincy, Mass. Government authorities in theUnited States have refused to allow operation ofthe Staten Island terminal, which is now a com-plete financial loss unless the facilities are usedto store LNG for peak shaving. One of the shipshas subsequently been incorporated into anLNG trade from Indonesia to Japan while theother is unemployed, and thus far a loss to itsowners.

As mentioned earlier, some supply contractscontain explicit provisions for periodic reviewof prices, while others may have to be renegoti-ated under changed circumstances because theparties are in separate countries not governedby a common legal jurisdiction. In either case,although producers and purchasers both facesubstantial incentives to come to terms, negotia-tions could conceivably break down after the fa-cilities have been built and put into operation.

Project failure can also occur if facilities aredestroyed by natural causes, civil strife, or war,or if a major change in attitude within one coun-try causes project termination. For example, theKhomeini government in Iran is reported tohave canceled the IGAT 11 project scheduled tosell gas to the U. S. S. R., which would in turn sellgas to Czechoslovakia and gas companies in Ger-many, France, and Austria, beginning in 1981.Unfortunately, Iran canceled the project afterconsiderable investment in pipelines.


Project interruption or delay

Technical problems can delay or interruptLNG projects. Normally, when any complexplant starts up, problems arise. In LNG liquefac-tion plants, typical problems include clogging ofcooling water intakes by sand, seaweed, jelly-fish, or debris; failures in the blading in tur-bines, compressors, or pumps; fouling of heatexchangers on the water side due to buildup ofalgae; clogging, possibly by ice, of spray rings,for cooling the storage tanks; or bearing failuresin pumps. 19

However, other technical problems can causelonger delays or force cessation of LNG projects,causing severe financial impacts. For example,when the Libyan LNG plant was first started in1969, several problems arose in the pipelinebringing gas to the liquefaction plant, resultingin additional delay of over a year. Libya also in-terrupted the LNG project by government edictwhen negotiations over prices broke down.

In another case, the main heat exchangers inthe liquefaction train of the Skikda, Algeria,plant were shut down for extended periods dueto a combination of mechanical wear caused byvibration, which ruptured heat exchangertubes, and corrosion caused by mercury in theheat exchangers.

Delays have also occurred in ship construc-tion, and at least two yards have been unable tofabricate the LNG containment systems onschedule. However, a surplus of LNG ships,some in layup and awaiting employment, couldcompensate for delays in the construction ofnew ones.

Cost overrun

The cost of any project, especially during aninflationary period, can be greater than antici-pated. If construction costs rise too fast and fi-nancing for the overrun cannot be found, theproject will not be completed and all invest-

ments are lost. More likely, the project will becompleted with added financing, but the costoverrun must ulimately be borne by someone.Operating costs can also exceed expectationsand cause either the sales price to rise or theearnings of project sponsors to fall.

Market uncertainties

Since the projects involve long-term contractsand investments, losses will result if the LNGshould not be marketable up to 20 years in thefuture at prices that cover the sponsors’ costs.Factors that contribute to uncertainty in thisarea are the possibility of increased domesticfuel production and conservation, the unpre-dictability of long-term economic growth rates,the unknown future course of world oil prices,possible changes in regulatory policy, and theoutcome of any supply contract renegotiations.

Who bears the financial risk?

The costs of project failure, operating or capi-tal cost overruns, supply interruption or reduc-tion, damage to facilities, or adverse govern-mental decisions, are borne by the various par-ties in an LNG project. These parties are theowners of the liquefaction plant, the ships, andthe receiving/regasification terminals; the lend-ers to the project; the guarantors of financing;the various governments who tax or otherwisereceive revenues from the project; the insurersof the facilities; and the consumers of the regasi-fied LNG. The distribution of risk is determinedby contracts among the parties; the financingagreements binding owners, lenders, and thosewho guarantee financing; the tariffs establishedby regulatory agencies; tax codes; and insur-ance agreements. The precise way in whichthese contractual instruments and tariffs dividethe risk will vary according to negotiations andregulatory decisions which take place when theproject is financed.

In order to reduce the capital costs of the proj-ect and thus increase profits and reduce the fi-nal price of LNG to the consumer, a substantialportion of the investments need to be financedby lenders who provide long-term loans at mod-est rates of interest. Lenders include banks, in-surance companies, and governmental finance


organizations such as MarAd and Eximbank,and their need to minimize exposure stronglyshapes the risk distribution of an LNG project.Banks and insurance companies lend other peo-ples’ money, that of their depositors, other cred-itors, or policyholders, and they are obliged torepay in full. Therefore, lenders, especially pri-vate banks and insurance companies, want theirmoney back no matter what happens to the LNGproject. They insist on guarantees of loan repay-ment from creditworthy parties, who can begovernments, natural gas consumers through“all events” tariffs, * or large corporations withongoing businesses out side of the LNG project.

The owners of the project, who provide theequity capital, money which is their own, willtypically absorb more risk in an LNG projectthan lenders, provided the potential return isgreater by virtue of their so doing. If the returnto the owners is limited, as it is in the UnitedStates for regulated utilities, the owner will alsoseek limitations of risk. The balance of risk andreturn that the project owners will accept is in-fluenced by their own special circumstancesand by other investment opportunities whichare competing for the equity capital needed byan LNG project. Finally, as a genera] rule, thoseparties who control the LNG facilities usually as-sume risk on it.

The Pac Indonesia project—An example of risk distribution

The Pac Indonesia case illustrates how risk isdistributed among the parties in an LNG project.While this project is somewhat different fromearly U.S. LNG import projects, it represents anappropriate example, since it is the only projectto be approved under the new organization ofDOE. The estimates are current and typical ofrecent LNG proposals, and the contractual rela-tionships reflect more than a decade of experi-ence. However, the reader should remember

that the distribution of risk in any project is inpart determined by the environment at the timeof negotiations,

Although generally representative, the Pac In-donesia project is unique in sevreral ways. First,Indonesia currently is completely dependent onJapan as its primary LNG customer. Also, theCalifornia gas utilities would like to use thesame west coast terminal for an additional proj-ect to bring LNG from the Cook Inlet in Alaska,and the proposed American shiponmer’s haveexpressed reservations and have not yet com-mitted their resources. Finally, the revolution inIran and the subsequent rapid increase in worldcrude oil prices during the first half of 1979have sharply altered perceptions about worldoil availability and price from the time the proj-ect was first negotiated and approved by DOE/ERA.

LIQUEFACTION AND LOADING FACILITIES

Indonesian liquefaction and loading facilitiesare estimated to cost about $869 million repre-senting the largest single capital portion of theproject. Pertamina, the national oil company ofIndonesia, is the owner. Mobil Oil IndonesiaInc., owns the producing facilities and bears thefinancial risk associated with them.

The instruments that distribute the risk arethe contract for the sale and purchase of LNGbetween Pertamina and Pac Indonesia, and thesecurity agreement for financing the liquefac-tion and loading terminal facilities. Two parts ofthe sales contract are important: the take-or-payand the force majeure clauses. The take-or-payclause requires Pac Indonesia to pay for theLNG tendered at the annual contract amount,whether or not the LNG is taken. The burden ofmarketing the LNG is thereby placed on Pac In-donesia, which is in a position to control its riskin this area. However, the force majeure clauseis broad and provides for cessation of contractobligations for acts of God, industrial strife, orgovernmental decisions interrupting the opera-tion of Indonesian facilities, ships, or U.S. facili-ties. This clause places most of the risk for theinvestment in Indonesia on Indonesian inter-ests. Lenders to Pertamina for the liquefaction


trains and base terminal are protected from fi-nancial loss by a guarantee of repayment fromthe Indonesian Central Bank.

Table 39 summarizes the risks and their dis-tribution for the liquefaction trains and loadingterminal. The perils that can occur, the mecha-nisms by which risk is transferred, the criteriagoverning the transfer, the amount transferred,who pays, and a reference to the source of theinformation are included in the table.

Force majeure events are borne solely by Per-tamina and other Indonesian interests. If costoverruns occur in the construction of Indone-sian facilities, they are borne by Indonesian in-terests, since the LNG price is not based on cost.An academic exception is that if the price ofcrude oil falls, and if the United States reversesits inflation and enters a deflationary period,the f.o.b. price for LNG may fall below a mini-mum established to ensure repayment of thelenders to the Indonesian facilities. In such acase, U.S. consumers are guaranteeing repay-ment of lenders, a risk most observers of oilmarkets and price behavior in industrial nationsview as insignificant.

However, LNG contracts do transfer somerisk to the buyer, and through the tariff, ul-timately to the U.S. gas consumer. If the U.S.dollar falls on foreign exchange markets, theLNG price is adjusted to ensure that Indonesiarecovers real value for LNG relative to a marketbasket of currencies. Also, Pertamina will not

accept the risk that the buyer will desire forwhatever reason not to take future LNG. Thus,if the LNG becomes unmarketable, the risk offailure in marketing the LNG is transferred tothe buyer and will be passed on through the tar-iff to the gas consumers.

SHIPPING

The six ships for the Pac Indonesia project, tobe constructed in U.S. shipyards for delivery in1983-84, are estimated to cost approximately$155 million (1978 dollars) each, excluding theconstruction differential subsidy by MarAd. Atotal of approximately $93o million to be paid bytheir owners and lenders will thus be at risk.Three other foreign ships have already beenconstructed at costs that are unknown but esti-mated at an average of $100 million (1976 dol-lars) each. In this project, the ships will beowned by independent shipowners, Ogden Ma-rine and Zapata, who provide the equity financ-ing, and chartered to Pac Indonesia.

MarAd will guarantee loans of up to 75 per-cent of the yard cost of a U.S. ship if a construc-tion differential subsidy is provided, or up to87.5 percent of the yard cost if no constructiondifferential subsidy is involved. These loanguarantees are available only to ships built inAmerican yards, for American owners, to haulgoods in a trade that includes America. How-ever, MarAd may negotiate with the shipownerfor additional money beyond the 25-percentequity interest to help protect the Government

Table 39.—Distribution of Financial Risk for Liquefaction and Loading Facilities of the Pac Indonesia Project

Risk transferEvent mechanism Criteria Amount Paid by References

1. Supply reduction or Sales contract Force majeure All Indonesian LNG salesinterruption interests contract

2. Project failure before or Force majeure All Indonesian LNG salesafter startup, liquefaction, interests contractor terminal problems

3. Cost overrun on Indonesian None. Price not If price drops, All Indonesian LNG salesfacilities cost-based minimum bill interests contract

4. Ship unavailable—no fault Sales contract, Not force majeure Difference in Shipowner limited LNG salesof Pac Indonesia, e.g., delay take-or-pay, LNG price when to 10% of capital contract,in ship construction charter hire, quantities made cost remaining charter hire

minimum bill customers5. Dollar depreciation in Tariff Automatic Ft~l U.S. consumers ERA 2, P.15

foreign exchange markets

aERA refers to an Opinion of the U.S. Department of Energy/Economic Regulatory Administration


guarantee. For example, MarAd required U.S.shipowners (Lochmar) in the Trunkline LNGproject to finance the 25-percent equity portionof the ships and to put up initial working capitalequivalent to 1 M years’ operation to provideadded protection in case the ships are not em-ployed immediately after they have been deliv-ered. Marine insurance covers losses from sink-ings, collisions, acts of God, and hull and ma-chinery failures.

Pending a resolution of who bears the finan-cial risk, the proposed American shipowners forPac Indonesia have not yet committed them-selves to providing the ships. If the proposedowners decide not to provide the shipping, Japa-nese firms are expected to do so. However, inthis discussion it is assumed that the Americanfirms will build the ships in American ship-yards, using a construction differential subsidy.

The risk of failure of other parts of the proj-ect, which would leave ships unemployed, arecontrolled by the charter arrangements be-tween the shipowners and Pac Indonesia, andby the U.S. tariff which governs how Pac Indo-nesia passes on its costs to consumers of naturalgas. Pac Indonesia has signed time charters for20 years with the various owners, under whichthe shipowner guarantees to deliver a ship ofspecified speed, fuel consumption, and boil-offrates. Pac Indonesia pays for the capital, main-tenance, and the operating costs of the ship.These payments are reduced if the ship does notmeet technical requirements or is not available,and the shipowner also assumes all costs whenthe ship is not available for service and the faultis not Pac Indonesia’s.

Through the time charter mechanism, manyof the risks are transferred to Pac Indonesia.However, the arrangements also specify thatPac Indonesia need not assume the risks if theyare passed on to the LNG consumer via an ap-proved tariff for foreign ships, or in the case ofa fall in the foreign exchange rate, offset by acurrency adjustment clause.

ERA and the administrative law judge in theiropinion and initial decision recommended a“minimum bill” provision of the tariff by whichPac Indonesia would charge Southern California

Gas and Pacific Gas & Electric for the LNG. Thisprovision specifies the costs that can be passedon even if the full amount of gas is not flowing,and includes the time charters or other arrange-ments for ocean transportation. In this way, thecost of shipping automatically passes to the gasconsumer after the project has begun,

Table 40 shows the sources of risk, the con-tractual instruments for distributing them, theamounts, and who pays. As the table shows, aslong as the project is operating and the ships areavailable, the charter-hire agreement and thetariff pass all costs of supply interruption or re-duction, increases in operating costs, or reduc-tion in the value of the dollar relative to a mar-ket basket of currencies, on to the gas consumervia minimum bill provisions. However, forevents that occur before the gas begins to flow,such as cost overruns on U.S. ships during con-struction, or project failure before or afterstartup, passthrough of costs to the consumer isnot automatic. In the latter case, FERC will de-cide if and to what extent gas consumers payafter an appeal under section 4 of the NaturalGas Act. Thus, what costs the consumer and theshipowner assume will be determined by an ad-ministrative ruling following an evidentiaryhearing to determine, for example, whethercost overruns were “prudent. ”

The proposed U.S. shipowners in the Pac In-donesia project have complained that they bearundue risk if they must depend on the outcomeof an administrative appeal in the event of costoverruns or project failure. The shipowners(ERA Decision No. 6, p. 9) have argued that un-less the customers are required through the tar-iff to pay for the charter obligations in the eventof project failure, MarAd title XI financing willnot be available. However, ERA and the admin-istrative law judge did not find sufficient evi-dence to support this claim. This finding is sup-ported by the fact that ship financing and con-struction for the Trunkline LNG project is pro-ceeding with MarAd loan guarantees. However,the interests, objectives, and perceptions of theproposed Pac Indonesia shipowners, which areotherwise independent of the project, may bedifferent from those of their Trunkline counter-parts; Panhandle Eastern Pipeline Company, the

.


Table 40.—Distribution of Financial Risk for Ships of the Pac Indonesia Project

Risk transferEvent mechanism Criteria Amount Paid bv References

1. Supply reduction or Minimum billinterruption

2. Project failure before or Sec. 4 type fillingafter startup, liquefaction,

Sec. 4 hearing

Charter contract

Tariff

Tariff

DOE review, tariff

Charter contract,minimum bill

-Automatic

Facts surroundingproject failure

Prudent

Foreign ships Customers

Charter fee, minus Customers

ERA 6, p. 8

parent of Trunkline LNG; General DynamicsCorporation, builder of the ships; and MooreMcCormack Bulk Transport, Inc., the operatorof the ships.

RECEIVING/REGASIFICATION TERMINALThe import terminal and regasification facili-

ties, with an associated pipeline to move the re-vaporized natural gas to existing pipelines, rep-resent the smallest of the investments in an LNGproject but are the ones on U.S. soil. In the caseof the Point Conception plan, these facilitieswould cost approximately $700 million (1978dollars) and could be used as an import terminalfor about 560 to 600 MMcf/d of Indonesian LNGwith remaining capacity for an Alaskan LNGproject of approximately 350 MMcf/d. In addi-tion, the Point Conception site is intended tostore LNG for peak shaving.

Approximately 75 percent of the costs of theterminal are expected to be financed by debt,primarily from banks or insurance companies,and the rest by equity capital. The distributionof risk is determined by the tariff and the secu-rity arrangements between the lenders andowners of the terminal as shown in table 41.

In its decisions, DOE/ERA very clearly distin-guishes between risk before startup and riskthat might occur after the gas is flowing.

Before project startup, ERA requires a section4 filing with FERC in order to determine whatcosts would be passed on to consumers. In thecase of a cost overrun, ERA suggests that onlyprudent costs be passed on to customers, andthat shareholders bear imprudent costs. Forproject failure before startup, no explicit cri-teria for passing on costs have been established.The ERA decision clearly states that the creditbefore completion should be provided by pri-vate creditworthy parties, who guarantee loans.Elsewhere, the FERC staff and others argue thatif the project fails, in no case should the equitycosts be passed on to consumers.

ERA recommends a minimum bill portion ofthe tariff providing that after gas first flows, thedebt portion of the terminals and other cost bepassed on automatically to gas consumers evenif the project fails. In the event of failure, supplyreduction, or interruption after project comple-tion, the ERA opinion (in contrast to the Trunk-line LNG decision) will also consider possible re-


Table 41 .—Distribution of Financial Risk at the Receiving/Regasification Terminal of the Pac Indonesia Project

Risk transferEvent mechanism Criteria

1. Supply reduction or a. Minimum bill Automaticinterruption

b. Sec. 4 type Extraordinaryproceedings circumstances

2. Project failure beforestartup

3. Cost overrun

4. Ships unavailable—nofault of Pac Indonesia orTerminal Associates

5. Project failure afterstartup

Sec. 4 filing Circumstances

Sec. 4 filing Prudent

Minimum bill is Automatic90% deliveries

Minimum bill AutomaticSec. 4 filing Circumstances

Amount Paid by

a. When less than Customer900/. delivered,no return of oron equity on un-delivered vol-umes. Recoveryof other allowedcosts.

b. Pro rata return Shareholdersof and on equtiy and perhaps

customersAs determined Customers

Allowed Customers,Disallowed ShareholdersCosts and equity Customers,on deliveries. Pro Shareholdersrata equity

Non-equity, CustomersEquity Shareholders

and customers

References

ID p. 81ERA 6, p. 13ERA 1, p. 30

ERA 1, p. 32

ERA 1, PP.32-33ERA 6, p. 11ERA 6, p. 16-17

ERA 1, p. 30ERA 6, pp. 11-14

ERA 1, P.33ERA 6, p. 13

turn of equity costs in the terminal regasifica-tion facilities in a section 4 proceeding if extra-ordinary circumstances can be shown.

SUMMARY OF RISK DISTRIBUTION

The distribution of risk in an LNG project, asexemplified by Pac Indonesia is as follows:.

Financial risk of the producing country facili-ties: ($869 million, 1976 dollars)

● Most risks are borne by the owner of theliquefaction facility, Pertamina, the gasproducer, Mobil, or the Indonesian Govern-ment through loan guarantees.

● Risk of the marketability of the LNG is bornb-y the U.S. gas companies and gas con-sumers.

Financial risks of shipping: ($1,230 million,1978 dollars)

● Insurance companies take normal shippingrisks such as damage to the ship, ground-ing, storm, etc. Gas consumers ultimatelypay the insurance premium.

● While the project is in operation, gas con-sumers bear costs if LNG flow is inter-rupted or reduced.

●

●

●

If the project fails, the shipowner may bearthe risk, at least on his equity in the ship, al-though FERC, after evidentiary hearings,may pass on some costs to the gas consum-er.Cost overruns may be borne by the ship-owner, unless after evidentiary hearingsFERC decides to pass on prudent costs tothe gas consumer.If all else fails, the lenders for U.S. shipswith financing guaranteed by MarAd re-ceive payment from the MarAd FederalShip Financing Fund, and if that is ex-hausted, from the U.S. Treasury.

Receiving/regasification terminal: ($437 mil-lion, 1977 dollars)

●

●

Loss due to project failure before gas flowsmay be fully borne by shareholders, unlessFERC, after an evidentiary hearing, decidesto pass some or all costs on to gas consum-ers.After gas flows, the non-equity costs of theterminal are borne by gas consumers.Shareholders bear risk of loss of equity andreturn on it in proportion to the reductionof LNG flows except that FERC may pass on

106 . The Future of Liquefied Natural Gas Imports

equity costs to consumers in extraordinarycircumstances.

● Cost overruns may be borne by sharehold-ers unless FERC, in an evidentiary hearing,passes on prudent costs to gas consumers.

Risk of LNG embargo by the producingcountry

Four of the six largest actual or potential ex-porters of natural gas from the Eastern Hemi-sphere—Algeria, Iran, Indonesia, and Nigeria—are members of OPEC. The U.S.S.R. is an adver-sary superpower. Only the last, Australia, is amember of OECD. As oil exporters, OPEC mem-bers have demonstrated their readiness to im-pose increases in price at short notice on exist-ing contract terms. Some of them, also, haveembargoed crude exports for political reasons.

On the other hand, the supplier as well as thepurchaser experiences dependence on LNGtrade and faces incentives not to interrupt ship-ments. LNG exports characteristically involvesubstantially greater capital investment than ex-ports of oil of comparable energy content. Con-tracts for 15 to 25 years of deliveries after atleast 5 years of negotiation and plant construc-tion, are necessary in order to amortize hugeinitial financial outlays. A large proportion ofthis investment, in gathering and trunk pipe-lines, liquefaction, plants, and terminals, is inthe exporting country, and in all cases so far,host governments have participated in the fi-nancing. If performance under a long-term con-tract were interrupted, alternative exports thatwould maintain revenues to pay capital chargeswould be very hard to arrange. The projects aretechnically integrated, and the only mobile capi-tal involved, the cryogenic tankers, will often beunder the effective control of foreign joint-ven-ture partners or import customers, and untilnow, no “merchant trade” in LNG has devel-oped. (Some LNG tankers have been built specu-latively, and as the international trade expands,a fringe of uncommitted tonnage will no doubtbecome available for the occasional balancingtransaction between customers with terminals.)Furthermore, selling the gas in the exporter’sdomestic market would be disadvantageous, be-cause local customers do not use LNG and may

not be located close to the pipelines leading toliquefaction terminals, and export volumes aregenerally surplus over domestic consumptionanyway,

These characteristics of present internationalgas trade seem to lock all parties to an LNG sup-ply contract into a closed economic loop. Thecosts of interruption, and of insurance against it(technical as well as financial), will be heavy.The resource is not lost (in the case of nonasso-ciated gas), but all parties will share all the costof downtime on a large accumulation of costly,dedicated capital.

This generalization in no sense precludes ar-guments over price while a contract is in force.It simply increases the pressure on all parties tosettle short of interrupting the gas flow. Also,20-year contracts in an environment of rapid in-flation and rising real prices for alternativefuels require effective escalation and reviewclauses.

In extreme circumstances, the high cost of in-terruption will not necessarily prevent gasbeing cut off for political purposes. one of thefew cases on record of an LNG contract’s actual-ly being interrupted was Libya’s action againstthe EXXON trade to Italy and Spain, which in-deed involved associated gas and was thus morecostly for both sides. * That arbitrary action mayhave been at once commercially and politicallymotivated. Although interrupting a gas opera-tion for political purposes is demonstrably morecostly than interrupting oil exports, it is none-theless possible.

The risk of supply curtailment can be reducedthrough policies that require the “exchange ofeconomic hostages, ” as suggested by Resourcesfor the Future. Under such a policy, “the pro-ducing country would own liquefaction facili-ties and tankers, and would finance them withborrowings other than from U.S. parties or gov-ernment entities such as the Export-ImportBank. ’)zo

4 Project Structure, Cost, and Financing

Documents

Transcript of 4 Project Structure, Cost, and Financing