- Email: [email protected]
The economics of natural gas development
Energy Vol. 10, No. 2, pp. 237-248. Printed in Great Britain
1985
0360-5442/85 53.00 + Ml 0 1985 Pergamon Press Ltd.
THE ECONOMICS OF NATURAL GAS DEVELOPMENT? BOURCIER, D. JULIUS, P. MOULIN Energy Department, World Bank, Washington,
P.
AND K. PALMER D.C. 20433, U.S.A.
(Received October 1983) Abstract-There is significant potential for natural gas to meet a growing share of developing countries’ energy demands. The constraints to rapid gas development are not related to supply potential or cost but rather to the country-specific problem of evolving a gas-investment framework, in which producer and consumer prices play a central role, that is conducive to rapid and matched growth of gas supply and demand. The paper considers three linked questions. First, what gas-utilization patterns can be expected in developing countries and how will they differ from those in developed countries? Second, what principles should determine the price of gas for local use and how can they be applied in practice? Third, how can exploration agreements provide incentives for investors to explore in gas-prone areas? These questions are explored drawing on World Bank experience in lending for gas projects in developing countries.
I.
INTRODUCTION
Background Natural gas reserves have been discovered in about 50 developing countries including 30 which are oil importing. In 1980, natural gas consumption averaged 7% of total commercial energy use in these countries (with wide variations among countries), compared with an almost 20% share in developed countries. A small number of developing countries have successfully developed a national gas industry (Argentina, Algeria, Pakistan, Iran, Venezuela, and Mexico), and a number of others are in the early stages of gas development (e.g. Egypt, Bangladesh, Bolivia, Nigeria, India). Excluding OPEC countries, the proven reserves to production ratio for developing countries is about 100 yrs, but this value understates the resource availability since many basins have not been properly explored so far. Clearly, then, resource availability will not be a constraint on rapid gas development in developing countries over the next decade. Recent experience with gas projects in a number of developing countries has also demonstrated their strong economic viability. Development and transmission costs have been lower and domestic gas demand higher and more diverse than had originally been expected.’ The full cost of developing and distributing gas to the consumer in these countries has rarely exceeded $ 2.00/MMBTU (MMBTU = lo6 Btu) compared with a delivered cost of competing oil products of more than twice that amount. Thus, for a developing country with indigenous gas resources and large oil-import bills, gas development is a high priority claim on scarce public investment funds. The constraints to rapid gas development in developing countries are typically not based on supply potential or cost. Rather, they are the complex and country-specific problems of evolving a gas policy framework, in which producer and consumer prices play a central role, that is conducive to rapid and matched growth of supply and demand. Such a framework must satisfy the objectives of the country in using its gas to best advantage, while also attracting foreign capital by safeguarding the interests of investors and lenders. Role of the World Bank The World Bank has been involved in natural gas development for many years. Early projects financed pipelines in Bolivia (for export to Argentina), Pakistan, Tunisia and Yugoslavia, as well as part of the first LNG plant at Arzew in Algeria. Since 1977 when t The views expressed World Bank.
in this paper are those of the authors
237
and do not necessarily
reflect the position
of the
P. BOURCIER
238
the Bank first began lending for hydrocarbon production, the number and variety of gas projects has grown rapidly. Over the past 5 yrs, more than $ 1 X lo9 has been lent for projects ranging from delineation drilling to city gas distribution in 10 countries. In the future, gas projects are likely to constitute a growing share of total Bank energy lending. This is partly because of the greater difficulty of attracting foreign equity capital to gas than to oil projects. Nonetheless, a primary objective of World Bank involvement in gas projects is to catalyze additional sources of finance, both equity and commercial cofinancing, for gas development and transmission. In Sec. V of this paper we discuss examples of how this has been done in practice. As noted already, however, one of the key requirements for attracting external finance is a gas policy framework that is reponsive to the interests of producers and users. Through involvement in project preparation, the Bank assists the government in taking a sector-wide view of supply and demand, in identifying and matching priority supply and use investments and in developing an appropriate pricing framework. The Bank also contributes to strengthening public sector institutions so that they can meet their commitments, and to the evolution of a feasible financing plan for gas sector investments. This is a role that is difficult for private investors and financiers to assume. They generally lack the long-term and widespread involvement with the country that the Bank has developed through its policy dialogue with the government on both macroeconomic and sector issues. Energy sector policies have the best chance of effective implementation when they are consistent with the country’s broader approach toward achieving its economic and social aims. In this paper we present some of the policy approaches that have evolved through the World Bank’s experience with natural gas development. These are not meant to be interpreted as universal remedies to the diverse problems of gas strategy in developing countries. However, we have tried to focus on common problem areas and to present general approaches that can be used to tailor specific solutions. In particular, we consider three linked questions. First, what gas utilization patterns can be expected in developing countries and how will they differ from developed countries? Second, what principles should determine the pricing of gas for local use? And third, how can exploration agreements ensure incentives to explore for and develop natural gas resources? Finally, in Sec. V, we briefly present the examples of Thailand and Egypt to illustrate more specifically how some of the problems in financing gas development have been overcome. II.
GAS
UTILIZATION
PATTERNS
IN
DEVELOPING
COUNTRIES
As noted above, there is good reason to believe that a large number of developing countries stand on the threshold of major programs of gas development. The bulk of this development will be geared not to producing gas as an export product, either directly or indirectly, but rather to the replacement of other fuels and feedstocks used to meet domestic demand. The electric power sector will remain the single largest consumer, followed by industry. The amount of gas consumed as fuel by those two sectors, about 45% of the total, is expected to be at least 50% larger than the total gas exported by developing countries as LNG. The “chemical” or feedstock uses of gas will comprise most of the remainder, and account for about 20% of the total. The value of gas to a developing country in a particular use may be quite different from what it would be in the same use in a highly industrialized country. For example, studies undertaken in several developing countries indicate that the net-back? value of gas used as fuel in the power sector is frequently higher than that of gas used as a feedstock for ammonia/urea production. This results primarily because of three features of the respective power and fertilizer markets. First, the construction and/or conversion time required to generate power from gas is much less than that required to commission t The net-back value of gas for a particular use can be thought of as the unit price of gas that would cause the project to just break even over its lifetime (including an appropriate return on the capital employed). It is calculated according to the following formula: Unit Gas Value =
NPV of project cash flow excluding Sum of PV of gas consumed
cost of gas
(in physical
units)
The economics of natural gas development
239
a new plant to produce fertilizer. This means that present value comparisons of the two alternatives will show the benefits materializing sooner by using the gas for power generation. Second, the capital requirements for ammonia/urea plant construction are much larger per unit of gas consumed than they are for either a gas turbine or combined cycle generating plant. In developing countries where capital is scarce, its opportunity cost is often as high as lo- 12% in real terms before accounting for inflation, compared to around 3-5% in developed countries. The relatively greater capital intensity of fertilizer production also means that returns are extremely sensitive to capacity utilization rates. A recent study by an authority on the fertilizer industry noted that under certain conditions, “the effect of operating at 70% rather than 90% (capacity utilization) is equivalent to having to pay an increased gas price of $ 2.00 X lo6 Btu”.2 The third reason that power often yields a higher return to gas than fertilizer in developing countries is that fertilizer is clearly a traded commodity, and its cost of production in the successful exporting countries has generally been based on very low gas input prices. This means that it is possible for many developing countries (depending on location and market size) to import urea at a price that already reflects low gas costs. The savings in ocean freight made possible by domestic production is often offset by higher site development costs. The increasingly competitive international market in fertilizer also introduces a large element of market risk in any developing country investment that is export oriented. For power, on the other hand, the alternative to gas is not direct import but use of alternative inputs such as fuel oil and coal. Many countries have already exploited their inexpensive hydropotential and lack both indigenous coal resources and the port and railway infrastructure necessary to import coal. Thus, the full value of gas as a fuel oil alternative is often realized in its use in the power sector. This is not to assert that in developing countries the power versus fertilizer comparison will always favor power. Rather, the point is that accepted “rules of thumb” derived from developed country experience for assigning priorities to gas uses may not apply in developing countries. Preliminary results from Bank studies on the value of gas in alternative uses indicate that the ranking of uses is highly country-specific and depends strongly on the pattern of past investment in gas-using plants. The design of supply systems may also differ in developing countries, especially those with surplus gas. The low labor costs and low marginal value of such gas will often make it advantageous to design systems with lower capital costs, lower energy efficiency and more labor intensive construction, operation and maintenance. This will, for example, influence the choice between pipeline looping and increased line compression. On the demand side too, a low marginal value of gas may make it economic to develop gas-using technologies which are not justified in developed countries. For example, in the transportation sector, because of the high value of gas in developed countries, there has been only limited conversion of fleet vehicles to compressed natural gas (CNG) and liquified natural gas (LNG). In gas surplus developing countries (e.g. Nigeria, Algeria, Bangladesh), however, Bank studies indicate that such vehicle conversions can show an acceptable economic rate of return.3 Thus, planning an appropriate gas strategy for a particular country and formulating pricing recommendations must generally be preceeded by a careful analysis of the demand potential and priority uses of gas in that country. These, in turn, will be highly dependent on such country-specific factors as the future fuel mix in the power sector, the growth of the agricultural sector requirements for fertilizer and even the cooking preferences of urban households for kerosene or charcoal. Yet, very often prefeasibility studies for gas development place inadequate emphasis on demand analysis and focus almost exclusively on the supply side questions of reserve size and potential deliverability, questions of obvious importance but subject to considerable uncertainty and providing only part of the information required in allocating initial gas supplies. III.
GAS
PRICING
IN LOCAL
MARKETS
An appropriate strategy for natural gas development should have as its main objective the maximization of net benefits to the country from the use of its exhaustible gas resources. This objective has three important dimensions, each of which implies certain
240
P. EOJRCIER
pricing principles. First, there must be the incentive to promote efficient use of the gas. Gas prices must be neither so high as to inhibit consumption (especially where the users must incur some cost to switch from other fuels), nor so low as to encourage wasteful use. Secondly, there must be adequate incentive to explore for and produce the gas. Particularly in cases where the government may be able to attract foreign capital to assist in gas development, the provision of an appropriate pricing and contractual framework is essential. (This topic is taken up in the next section.) Finally, the growth rates of both supply and demand for gas should be rapid and matched up to the level where full development has been reached. As discussed below, the basic principle that facilitates the achievement of all three objectives is that both consumer and producer prices should be set near the marginal opportunity cost of the gas. Excess producers’ or consumers’ surplus (excess profits) should be captured through profit taxation. In practice, this approach is complicated by uncertainties affecting reserve size and the growth rate of the market. Before considering these complications, however, we focus on the problem of determining the opportunity cost of gas under assumed conditions of known (or predictable) supply and demand. The meaning of opportunity cost The opportunity cost for gas, or any other commodity, can be thought of as the price that will equate demand and supply. If the good is internationally traded, then the relevant import supply and export demand functions must be included in the calculations. An example of this situation is shown in Fig. 1. If the good were not traded, its demand and supply curves would be those labeled “domestic demand” and “domestic supply.” Its opportunity cost to the country would be P NTand the appropriate quantity to produce would be QNT. Once there is an international market in the good, however, the relevant demand and supply curves must take into account the import and export possibilities. If the good can be imported at a price P, and exported at P, (where the difference between P, and P, represents the freight, insurance and handling cost of trade), then the relevant demand and supply curves become the linked one labeled “traded supply” and “traded demand.” Their intersection is at the price P, where the quantity Q, will be produced, QC will be consumed and the difference (Qp - QC) will be exported. In this case, the availability of an international market means both that more should be produced and
DomeStIC SUPPlY
Tradable SUPPlY
Tradable Demand
I
Qc Fig. 1. Supply and demand
I
QNT for traded
I I
QP
c Quontiiy
and nontraded
goods.’
The economics
of natural
gas development
241
also that a higher price should be charged to domestic consumers than if there were no export market for the good. The net gain to the country from producing and exporting the amount Q, - Q, is greater than the net loss to the country of producing and consuming only QNT at the lower price. This simplified diagram demonstrates the importance of the tradablefnontradable distinction in determining opportunity costs. Natural gas, of course, is generally not a commodity that is directly traded by developing countries. However, the distinction is still relevant as long as the gas is used domestically to substitute for another commodity (such as fuel oil) that is tradable. Gas only becomes nontradable in the economic sense when, at the margin, additional supplies that could be produced can no longer find any local markets where they would be replacing traded goods. The demand curve
An example may help to clarify this point. Figure 2 shows a demand curve for gas that is derived from composite data from two studies on the unit value of gas in various possible uses for a middle-income developing country. The length of each “step” represents the net-back value of gas derived from the export or import prices of the goods it is used to produce. For this particular country the highest value uses of gas are for peak power generation (where it essentially replaces diesel oil), household distribution (where it replaces LPG and kerosene) and methanol. The total amount of gas that can be consumed for those purposes, however, represents less than 5% of the potential market even excluding LNG. The fuel oil and coal that can be replaced in the power and industrial sectors clearly constitute the bulk of the market. Production of fertilizer and steel based on gas is sized to replace all imports of those commodities and generate a surplus for potential exports until around the year 2000 when domestic demands are projected to be large enough to absorb the full output. The LNG net-back value is based on a 2 X lo6 ton/yr facility which was judged to be the maximum amount which the country could reasonably expect to sell. For these reasons, if one were to visualize how the curve in Fig. 2 would shift over time, certain steps would grow longer (i.e., those based on domestic demand such as household distribution and fuel oil replacement) while others (i.e. fertilizer, LNG) would probably remain unchanged. The supply curve
Turning now to the gas supply picture, Fig. 3 illustrates a stepped cost function where the length of each step represents an amount of sustained production that could be
A
I
1.oco Consumption
of Gas (MMCFD)
Fig. 2. 1990 gas demand.’
P. B~URCIER
242
costof Gas (SMMBTU)
Discovered
“Probable”
SUPPlY
SUPPIV
r-----0 l.MM
500 Supply
I.500
2.oca
of Gas (MMCFD)
Fig. 3. 1990 gas supply.’
delivered for the incremental cost plotted on the vertical axis. The first (lowest) step in this function shows an amount of 100 MMCFD (MMCFD = lo6 ft3/d) of onshore, associated gas available at an incremental cost of $0.20/106 Btu. The second and third steps show production of nonassociated gas from onshore and offshore fields, respectively, at progressively higher costs. The sum of these three steps gives the country’s projected 1990 gas supply based on today’s proven reserves. The dashed line, drawn as the fourth step, shows that the full cost of finding and producing an additional 100 X lo6 ft3/d from reserves presently classified as “probable” is estimated at $2.50/106 Btu. The supply curve shown in Fig. 3 is a simplified picture which abstracts from at least two important complications. First, it represents deliverable rather than potential supply, and therefore incorporates considerations of appropriate field depletion rates and possible infrastructure constraints. In actual practice, the steps of the curve would be less abrupt since it is usually possible to increase production somewhat through added compression or temporarily faster depletion, A second, and more important, qualification is that the costs shown in the curve do not include any component to represent the opportunity cost to the country of consuming its finite gas resources now rather than in the future. Estimating, and accounting for, this depletion premium or user cost is a complex but important task in countries where gas is likely to be in excess supply for a relatively short period.7 Having discussed the derivation of both demand and supply curves, we are now in a position to superimpose the two in order to determine the opportunity cost of gas to this country. Figure 4 shows the result. Based on potential production from proven reserves, 1990 gas availability would be sufficient to meet all of the uses down to and including steel production and about one-tenth of the coal substitution. This would indicate that the opportunity cost of gas would be derived from its value as a coal replacement, equivalent to about $2.00/106 Btu. At this value, the probable reserves would not be developed since the cost of that gas would be higher than the equivalent cost of coal while all of the higher valued uses for gas were already being served. If one visualizes these curves in 1995 or the year 2000, however, rightward shifts in the demand function would make it profitable to develop the probable reserves. In that case, the intersection t Work carried out in this area at the World Bank indicates that the depletion premium at the beginning a period of IO-15 yrs of supply surplus may still amount to 50-607’0 of the fuel oil equivalent price.
of
The economics of natural gas development
243
Price of Gas (S/MMBTU)
Discovered
“Probable”
SUPPlY
SUPPlY
a-
I I
-------
r-“-
o
p -----__
_____
j
L
-,J
1 A
-m-J
I
I
500
1,000 Quantity
I I.500
1 2.000
of Gas (MMCFD)
Fig. 4. 1990 opportunity cost of gas.’
of the supply and demand curves would be on a horizontal segment of the supply curve. This means that the opportunity cost of gas would be equivalent to its incremental supply cost (abstracting from the depletion premium issue mentioned above), and no coal substitution would take place. Problems with opportunity cost pricing While the economic advantages of opportunity cost pricing are clear and, as shown above, it is not an impossible empirical task to derive such prices, it must be admitted that few gas-consuming countries follow this approach. There are at least two reasons given which are examined below. First, pricing at the marginal opportunity cost means that intramarginal producers and consumers may reap large excess profits. For example, in Fig. 4 if gas is priced at $2.00/106 Btu, then the least-cost producer would earn (before tax) rents of $1.80/106 Btu. However, a profits tax is an efficient and practical device to ensure that these surpluses are returned to the rest of the economy. A uniform price combined with profits taxation simply parallels the universal practice for crude oil. (Further details of practical profits taxes are described in Ref. 4.) A second objection to opportunity cost pricing in some countries is that the prices of competing fuels may be subsidized (or, more rarely, taxed) at levels which would encourage uneconomic fuel choices. For example, if fuel oil were subsidized and sold at a price equivalent to $1 SO/ 1O6 Btu, then pricing gas at its opportunity cost of $2.0011 O6 Btu in the case shown in Fig. 4 would discourage fuel oil users from shifting to gas. Clearly, the best solution to this problem would be to remove the fuel oil subsidy but, if that is not immediately possible, the government may consider delinking the producer and consumer prices of gas in order to permit the consumer price to be competetive with that of fuel oil while retaining the producer’s incentive to explore and produce gas. IV.
NATURAL
GAS
EXPLORATION
INCENTIVES
Most exploration/production agreements do not include specific gas pricing provisions. Instead, there is often only a general provision that, in the event of a significant gas discovery, the investor and the government will negotiate a price. In many cases, the price negotiated under such agreements has been determined on a cost plus rate of return basis at a level adequate to justify incremental investments but insufficient to justify the
244
P. BOURCIER
initial high risk exploration. Given this experience, in many countries private oil companies have been unwilling to explore in gas-prone basins despite attractive geological potential. Providing a more explicit gas pricing provision which links the price to its market value is a necessary condition to induce greater exploration in these areas. The central problem in specifying a specific gas price prior to exploration is supply uncertainty. In advance of exploration, it is not possible to predict the size of future discoveries. If there is a discovery whose supply potential is large relative to domestic demand (but too small or costly for export) then the marginal value of gas in that country will be altered dramatically. This is illustrated in Fig. 5. Initially we assume domestic gas demand exceeds supply [Fig. 5(i)] with domestic gas production OQ, up to the point where marginal costs equal the price of the alternative input OP, [Fig. 5(i)]. We assume the supply-demand gap Q2Q1 is met by fuel oil. In Fig. 5(ii) we assume a significant gas discovery with costs of production below OP, which shifts the gas supply curve to the right to ABCE. Now, domestic gas supply completely displaces fuel oil, and the marginal value of gas is reduced to OP2 with production OQ . The diagram illustrates that whenever uncertainty about the supply curve of a nontraded good is large relative to domestic demand, then the expected price of that good is also very uncertain. Herein lies the difficulty of specifying explicit gas pricing provisions in exploration contracts. At the very least, however, such pricing provisions should establish the principles and methodology to be adopted in the event of a commercial gas discovery. Contractual language might specify only that the parties will agree, following discovery, on a price determined by reference to the marginal opportunity cost of the gas at the time. Although such a general provision is subject to interpretation, taken together with provisions for independent technical input and arbitration, this language has the merit of establishing the principle of opportunity cost pricing rather than cost plus pricing and hence altering investor expectations of the likely future price. It represents a considerable improvement over contracts which simply state the parties will negotiate a price without establishing any clear pricing guidelines, particularly in countries where there is a recent history of very low producer prices (e.g. Egypt and Pakistan). Another important step to promote exploration for gas in developing countries is to ensure contractors a right to export gas in excess of a specified national reserve, as was recently done in Egypt, and to receive the international price for those exports. This provides a clear, prior guide to the company that should it make a certain size discovery it will have access to those reserves for export. Although this type of national reserve provision needs to deal with problems of prorationing of export entitlements and independent audit of proven reserves, the case of Egypt indicates that solutions can be found. The main problem here is helping the government to determine the minimum necessary size of, or formula for determining, such a national reserve. Supply constrained countries
Developing countries can usefully be categorized into those which are gas supply constrained and those which are gas demand constrained. In the former case, a scarcity of supply relative to demand means that a predictable value can be placed on additional discoveries of gas on a delivered basis up to the limit of excess demand. In demandconstrained countries, domestic demand is small relative to proven gas reserves so the marginal value of gas in the ground approaches zero. Different approaches will clearly be required in the two types of countries to promote appropriate levels of exploration investment in each. In supply-constrained countries, guaranteed purchase provisions and gas buy-back provisions are two of the specific measures that have been used to promote additional exploration. The formula can be used whenever the government can predict with reasonable confidence the magnitude of the supply-demand gap in the medium term. Guaranteed price provisions in exploration agreements would provide a specific gas price formula for the next x million cubic feet per day of gas delivered where the quantity would be equal to the expected gas shortfall. For example, if a gas supply shortfall over the next decade in a country is expected to result in the conversion of a thermal power
The economics of natural gas development
245
m
8
‘C
a
246
P.
BOURCIER
generating plant from gas to fuel oil (or new plants are to be oil fired) then the gas price formula for a quantity of gas up to x X lo6 ft3/d would be linked to the price of fuel oil.? This pricing incentive (by transferring price risk to the government) would act as an important stimulus to exploration investment in gas-prone areas. The guaranteed price provisions would be combined with the profit sharing provisions applicable to oil to ensure that, for low cost, intramarginal producers, a large share of producers’ surplus was returned to the country. One problem with guaranteed price provisions is that they presume that newly discovered gas in a supply constrained situation should be developed. This is not necessarily the case where supply is constrained not by physical delivery limitations, but by the desire to maintain a certain production/reserves ratio. Often a new discovery will not be the least cost means of adding to supply; rather the discovery should be added to reserves for future use and the rate of production from existing fields with established infrastructure increased. But here, clearly, the national and investor interests diverge: For the investor there is no point in exploring just to add to national reserves. One solution, in this situation, is a buy-back provision where, if the investor discovers proven reserves in excess of some minimum size (say 1 X lOI ft3), and if the investor and government do not agree on a development program, the government undertakes to reimburse the contractor’s exploration costs in return for relinquishing the discovery. A sliding scale premium might be paid for larger reserves. Although reimbursement of costs would not, in itself, be sufficient to induce gas exploration, taken together with the potential for oil it would enhance the expected value of an area. The buy-back provisions would have to be limited to the quantity of gas which could be expected to have value to the country in excess of the cost of buy-back. It would also be necessary to specify in advance the reimbursable exploration costs. These should be limited to a particular program of wells in order to avoid creating an overincentive, at the margin, to drill additional wells merely to obtain reimbursement of sunk costs. Additional exploratory or appraisal wells by the investor beyond the agreed program would not be reimbursable. There would also be a need for independent audit of reserves. This type of approach, which has recently been adopted in Egypt, could be applied in a number of other supplyconstrained countries. The assurance of buy-back up to the level of desired national reserves combined with the right to export reserves in excess of that level would be a significant exploration incentive. Demand constrained countries Neither of these approaches work in demand-constrained countries where the absence of a predictable gas use means an ex ante pricing basis cannot be established. In demandconstrained countries, where proven gas reserves exceed requirements for domestic use but are too small or too expensive for export projects, the economic value of gas in the ground approaches zero. In this case, exploration contracts can only establish the principle of opportunity cost pricing. It may be argued that in demand-constrained countries there is no need to encourage gas exploration, but this is not necessarily the case in countries with no proven reserves and negligible domestic demand. It may be that only following a discovery can potential demand be converted into effective demand. However, since users will not negotiate deals just in case a discovery is made, the lack of certainty regarding the m.arket for gas is a significant deterrent to gas exploration. The fact that a discovery in such a country has a low economic value can attract footloose gas using industries such as fertilizer or methanol plants. In these situations countries should encourage such gas development projects whenever the estimated netback value of the gas in use exceeds the marginal investment costs of field development and transmission (plus any depletion margin that may be imputed to reserves in the ground which in these countries is likely to be very small). However, to attract these industries there must be an agreement over price. Potential users naturally seek to negotiate a low into-plant price arguing that a significant incentive is required to set up t The guaranteed price need not be 100% of the expected economic value of incremental supply. It could be 70% or even 50%, so long as the price remained above the expected marginal cost of exploration and development of new reserves by a significant margin so that producer incentives were retained.
The economics of natural gas development
247
in a developing country and that expected profitability is marginal. There is always considerable market risk confronting export oriented users. A low gas price, by reducing the downside risk, is often an important factor in persuading an investor to set up in one country rather than another. While the government wants to attract the user industry, the establishment of a low price runs the risk of transferring the major share of the benefits of gas ownership to the user should the project prove highly profitable. One solution to this predicament is a risk-sharing contract under which the gas price is conditional on the ex post profitability of the gas-using industry. This type of price arrangement contains two parts: (i) a base price, combined with a take-or-pay sales agreement set just high enough to cover the cash flow needs to amortize incremental field development and transmission costs (i.e., operating costs and debt service payments and possibly a return on equity); and (ii) a supplementary unit payment based on the ex post profitability of the gas-using industry that escalates with increasing ex post discounted cash flow (DCF) return on investment. This arrangement ensures that the government obtains a high share of the ex post rent from gas production while minimizing user front-end risk (and accelerating cash flow), thereby attracting the investment in the first place. Thus this approach can make the government competitive in attracting footloose industries to utilize gas that would otherwise have no value, without negotiating away permanently the national benefits. V. FINANCING
LDC
GAS
DEVELOPMENT:
TWO
EXAMPLES
Thailand and Egypt are two countries where significant progress in gas development has been made recently. The type of studies, technical assistance and external financing that was required in the two countries differed considerably, but the approach of dealing with project difficulties by developing a better perspective on broader energy sector issues was successfully applied by both countries. In both, the role of the World Bank was to assist the government in identifying the key sector issues and in formulating a program of studies and investments to address them. Thailand When the Government of Thailand approached the Bank in 1976, the situation was at a classical stand-off. The company exploring in the Gulf of Thailand had made an important discovery and was willing to proceed with appraisal drilling but only after a contract defining the conditions under which the gas would be sold to Thailand was agreed. The government, which had no previous experience in gas, was faced with a series of difficult decisions about production volumes and possible uses of gas as well as prices and financing arrangements. While the two parties had been holding discussions for some time, no agreement was in sight. At the request of the government a Bank mission visited Thailand and, jointly, with the Ministry of Mines, prepared a plan for a program of studies to provide the government with the basic data required to make the necessary decisions. Consultants were hired to evaluate the discovery and monitor future drilling, to evaluate the market for gas under various price scenarios and to advise the government during contract negotiations. The various studies concluded that there was a market for gas primarily as a substitute for fuel oil in power generation and that gas could be brought to the market under conditions which would be attractive to Thailand while providing an adequate return to the oil company. This led to a contract which enabled the company to proceed with delineation drilling. At this point, although considerable manpower had been expended, no financial commitment had been made. Once enough reserves were proven, the company financed field development while the government mobilized the financing required for the pipeline. In addition to the Bank loan of $ 107 X 106, this included export credits and commercial loans in excess of $ 300 X 106. The project was completed on time and well within the cost estimates.
In the case of Egypt, the government identified early in 1977 the potential economic benefits of substituting natural gas for liquid fuels in the domestic energy market, thereby
248
P. BOURCIER
increasing net oil exports. This recognition followed completion of a gas utilization study promoted and financed by the Bank. So far, the Bank has financed four gas projects in Egypt in support of the country’s national gas development plan. These projects deal respectively with the recuperation of flared gas in the Gulf of Suez, the construction of a domestic gas distribution system in Cairo, continued gas exploration in the Western Desert, and development of the offshore gas field of Abu Qir. Each of these projects included finance for technical and economic studies to address sector issues ranging from updating estimates of gas reserves to devising a better price structure for petroleum products. While relatively few cofinanciers have been involved in these projects, so far, we expect this to change now that the feasibility of gas development in Egypt has been established and that significant steps have been taken by the government to interest private oil companies in exploration in gas-prone areas, as well as to attract increased commercial financing for development and transmission investments. In general, World Bank financing covers only a fraction of total project costs. It is applied flexibly to finance items of less interest to commercial bankers. In particular, because of the longer maturities of Bank loans, they are often used to finance basic infrastructure which, while contributing substantially to the projects’ economic viability, may not generate immediate financial returns. Through the loan documentation, certain guarantees, cross default clauses for example, can also be provided to private bankers which may increase their willingness to participate in project financing. Recently the Bank has introduced several new financing instruments specifically designed to facilitate private sector lending to developing countries by taking advantage of the lower perceived lending risks resulting from association with the World Bank Group. Despite these new initiatives, the financing of gas development remains difficult in many countries. In the poorest developing countries despite attractive projects, the noncreditworthiness of the country precludes commercial financing. The Bank’s limited resources for soft loan financing (IDA) may prevent good projects from proceeding in precisely those countries which most need them.5 CONCLUSION
There is no doubting the substantial potential for natural gas to meet a growing share of developing countries’ energy demands. Our experience to date suggests that the optimal pattern of investment in the gas sector and appropriate gas pricing policies will vary widely among countries depending on their particular demand characteristics and supply endowments. Sector-wide supply and demand analyses will be a prerequisite for determining the marginal opportunity cost of gas which should be the determinant of the gas price. The absence of explicit gas pricing provisions in exploration agreements is a deterrent to exploration in gas-prone basins. Contracts should, as a minimum, establish the principle that gas will be treated in a similar way to oil with a uniform producer price based on the marginal opportunity cost and gas profits subject to similar profit sharing fiscal arrangements. In supply-constrained countries, more specific contract provisions can be devised to encourage exploration. The financing of gas projects in creditworthy countries does not create insuperable problems once an appropriate gas investment program is defined and pricing policies introduced which permit and encourage investors and financiers to commit themselves to the investment program. The central role of the World Bank is to assist governments in devising and implementing such an investment program. REFERENCES I. World Bank, The Energy Transition in Developing Countries (1983). 2. William F. Sheldrick, “The Effect of Energy and Investment Costs on Total Fertilizer Production Costs,” paper presented at ISMA Meeting, London, p. 6 (1981). 3. World Bank Energy Department, Energy Department Paper No. 4, “Alternative Fuels for Use in Internal Combustion Engines” (198 1). 4. K. F. Palmer, “Mineral Taxation Policies in Developing Countries: An Application of Resource Rent Tax”, IMF Staff Papers, Vol. 27, No. 3 (1980). 5. A. W. Clausen, “Address to the Board of Governors of the World Bank” (I 982).
1985
0360-5442/85 53.00 + Ml 0 1985 Pergamon Press Ltd.
THE ECONOMICS OF NATURAL GAS DEVELOPMENT? BOURCIER, D. JULIUS, P. MOULIN Energy Department, World Bank, Washington,
P.
AND K. PALMER D.C. 20433, U.S.A.
(Received October 1983) Abstract-There is significant potential for natural gas to meet a growing share of developing countries’ energy demands. The constraints to rapid gas development are not related to supply potential or cost but rather to the country-specific problem of evolving a gas-investment framework, in which producer and consumer prices play a central role, that is conducive to rapid and matched growth of gas supply and demand. The paper considers three linked questions. First, what gas-utilization patterns can be expected in developing countries and how will they differ from those in developed countries? Second, what principles should determine the price of gas for local use and how can they be applied in practice? Third, how can exploration agreements provide incentives for investors to explore in gas-prone areas? These questions are explored drawing on World Bank experience in lending for gas projects in developing countries.
I.
INTRODUCTION
Background Natural gas reserves have been discovered in about 50 developing countries including 30 which are oil importing. In 1980, natural gas consumption averaged 7% of total commercial energy use in these countries (with wide variations among countries), compared with an almost 20% share in developed countries. A small number of developing countries have successfully developed a national gas industry (Argentina, Algeria, Pakistan, Iran, Venezuela, and Mexico), and a number of others are in the early stages of gas development (e.g. Egypt, Bangladesh, Bolivia, Nigeria, India). Excluding OPEC countries, the proven reserves to production ratio for developing countries is about 100 yrs, but this value understates the resource availability since many basins have not been properly explored so far. Clearly, then, resource availability will not be a constraint on rapid gas development in developing countries over the next decade. Recent experience with gas projects in a number of developing countries has also demonstrated their strong economic viability. Development and transmission costs have been lower and domestic gas demand higher and more diverse than had originally been expected.’ The full cost of developing and distributing gas to the consumer in these countries has rarely exceeded $ 2.00/MMBTU (MMBTU = lo6 Btu) compared with a delivered cost of competing oil products of more than twice that amount. Thus, for a developing country with indigenous gas resources and large oil-import bills, gas development is a high priority claim on scarce public investment funds. The constraints to rapid gas development in developing countries are typically not based on supply potential or cost. Rather, they are the complex and country-specific problems of evolving a gas policy framework, in which producer and consumer prices play a central role, that is conducive to rapid and matched growth of supply and demand. Such a framework must satisfy the objectives of the country in using its gas to best advantage, while also attracting foreign capital by safeguarding the interests of investors and lenders. Role of the World Bank The World Bank has been involved in natural gas development for many years. Early projects financed pipelines in Bolivia (for export to Argentina), Pakistan, Tunisia and Yugoslavia, as well as part of the first LNG plant at Arzew in Algeria. Since 1977 when t The views expressed World Bank.
in this paper are those of the authors
237
and do not necessarily
reflect the position
of the
P. BOURCIER
238
the Bank first began lending for hydrocarbon production, the number and variety of gas projects has grown rapidly. Over the past 5 yrs, more than $ 1 X lo9 has been lent for projects ranging from delineation drilling to city gas distribution in 10 countries. In the future, gas projects are likely to constitute a growing share of total Bank energy lending. This is partly because of the greater difficulty of attracting foreign equity capital to gas than to oil projects. Nonetheless, a primary objective of World Bank involvement in gas projects is to catalyze additional sources of finance, both equity and commercial cofinancing, for gas development and transmission. In Sec. V of this paper we discuss examples of how this has been done in practice. As noted already, however, one of the key requirements for attracting external finance is a gas policy framework that is reponsive to the interests of producers and users. Through involvement in project preparation, the Bank assists the government in taking a sector-wide view of supply and demand, in identifying and matching priority supply and use investments and in developing an appropriate pricing framework. The Bank also contributes to strengthening public sector institutions so that they can meet their commitments, and to the evolution of a feasible financing plan for gas sector investments. This is a role that is difficult for private investors and financiers to assume. They generally lack the long-term and widespread involvement with the country that the Bank has developed through its policy dialogue with the government on both macroeconomic and sector issues. Energy sector policies have the best chance of effective implementation when they are consistent with the country’s broader approach toward achieving its economic and social aims. In this paper we present some of the policy approaches that have evolved through the World Bank’s experience with natural gas development. These are not meant to be interpreted as universal remedies to the diverse problems of gas strategy in developing countries. However, we have tried to focus on common problem areas and to present general approaches that can be used to tailor specific solutions. In particular, we consider three linked questions. First, what gas utilization patterns can be expected in developing countries and how will they differ from developed countries? Second, what principles should determine the pricing of gas for local use? And third, how can exploration agreements ensure incentives to explore for and develop natural gas resources? Finally, in Sec. V, we briefly present the examples of Thailand and Egypt to illustrate more specifically how some of the problems in financing gas development have been overcome. II.
GAS
UTILIZATION
PATTERNS
IN
DEVELOPING
COUNTRIES
As noted above, there is good reason to believe that a large number of developing countries stand on the threshold of major programs of gas development. The bulk of this development will be geared not to producing gas as an export product, either directly or indirectly, but rather to the replacement of other fuels and feedstocks used to meet domestic demand. The electric power sector will remain the single largest consumer, followed by industry. The amount of gas consumed as fuel by those two sectors, about 45% of the total, is expected to be at least 50% larger than the total gas exported by developing countries as LNG. The “chemical” or feedstock uses of gas will comprise most of the remainder, and account for about 20% of the total. The value of gas to a developing country in a particular use may be quite different from what it would be in the same use in a highly industrialized country. For example, studies undertaken in several developing countries indicate that the net-back? value of gas used as fuel in the power sector is frequently higher than that of gas used as a feedstock for ammonia/urea production. This results primarily because of three features of the respective power and fertilizer markets. First, the construction and/or conversion time required to generate power from gas is much less than that required to commission t The net-back value of gas for a particular use can be thought of as the unit price of gas that would cause the project to just break even over its lifetime (including an appropriate return on the capital employed). It is calculated according to the following formula: Unit Gas Value =
NPV of project cash flow excluding Sum of PV of gas consumed
cost of gas
(in physical
units)
The economics of natural gas development
239
a new plant to produce fertilizer. This means that present value comparisons of the two alternatives will show the benefits materializing sooner by using the gas for power generation. Second, the capital requirements for ammonia/urea plant construction are much larger per unit of gas consumed than they are for either a gas turbine or combined cycle generating plant. In developing countries where capital is scarce, its opportunity cost is often as high as lo- 12% in real terms before accounting for inflation, compared to around 3-5% in developed countries. The relatively greater capital intensity of fertilizer production also means that returns are extremely sensitive to capacity utilization rates. A recent study by an authority on the fertilizer industry noted that under certain conditions, “the effect of operating at 70% rather than 90% (capacity utilization) is equivalent to having to pay an increased gas price of $ 2.00 X lo6 Btu”.2 The third reason that power often yields a higher return to gas than fertilizer in developing countries is that fertilizer is clearly a traded commodity, and its cost of production in the successful exporting countries has generally been based on very low gas input prices. This means that it is possible for many developing countries (depending on location and market size) to import urea at a price that already reflects low gas costs. The savings in ocean freight made possible by domestic production is often offset by higher site development costs. The increasingly competitive international market in fertilizer also introduces a large element of market risk in any developing country investment that is export oriented. For power, on the other hand, the alternative to gas is not direct import but use of alternative inputs such as fuel oil and coal. Many countries have already exploited their inexpensive hydropotential and lack both indigenous coal resources and the port and railway infrastructure necessary to import coal. Thus, the full value of gas as a fuel oil alternative is often realized in its use in the power sector. This is not to assert that in developing countries the power versus fertilizer comparison will always favor power. Rather, the point is that accepted “rules of thumb” derived from developed country experience for assigning priorities to gas uses may not apply in developing countries. Preliminary results from Bank studies on the value of gas in alternative uses indicate that the ranking of uses is highly country-specific and depends strongly on the pattern of past investment in gas-using plants. The design of supply systems may also differ in developing countries, especially those with surplus gas. The low labor costs and low marginal value of such gas will often make it advantageous to design systems with lower capital costs, lower energy efficiency and more labor intensive construction, operation and maintenance. This will, for example, influence the choice between pipeline looping and increased line compression. On the demand side too, a low marginal value of gas may make it economic to develop gas-using technologies which are not justified in developed countries. For example, in the transportation sector, because of the high value of gas in developed countries, there has been only limited conversion of fleet vehicles to compressed natural gas (CNG) and liquified natural gas (LNG). In gas surplus developing countries (e.g. Nigeria, Algeria, Bangladesh), however, Bank studies indicate that such vehicle conversions can show an acceptable economic rate of return.3 Thus, planning an appropriate gas strategy for a particular country and formulating pricing recommendations must generally be preceeded by a careful analysis of the demand potential and priority uses of gas in that country. These, in turn, will be highly dependent on such country-specific factors as the future fuel mix in the power sector, the growth of the agricultural sector requirements for fertilizer and even the cooking preferences of urban households for kerosene or charcoal. Yet, very often prefeasibility studies for gas development place inadequate emphasis on demand analysis and focus almost exclusively on the supply side questions of reserve size and potential deliverability, questions of obvious importance but subject to considerable uncertainty and providing only part of the information required in allocating initial gas supplies. III.
GAS
PRICING
IN LOCAL
MARKETS
An appropriate strategy for natural gas development should have as its main objective the maximization of net benefits to the country from the use of its exhaustible gas resources. This objective has three important dimensions, each of which implies certain
240
P. EOJRCIER
pricing principles. First, there must be the incentive to promote efficient use of the gas. Gas prices must be neither so high as to inhibit consumption (especially where the users must incur some cost to switch from other fuels), nor so low as to encourage wasteful use. Secondly, there must be adequate incentive to explore for and produce the gas. Particularly in cases where the government may be able to attract foreign capital to assist in gas development, the provision of an appropriate pricing and contractual framework is essential. (This topic is taken up in the next section.) Finally, the growth rates of both supply and demand for gas should be rapid and matched up to the level where full development has been reached. As discussed below, the basic principle that facilitates the achievement of all three objectives is that both consumer and producer prices should be set near the marginal opportunity cost of the gas. Excess producers’ or consumers’ surplus (excess profits) should be captured through profit taxation. In practice, this approach is complicated by uncertainties affecting reserve size and the growth rate of the market. Before considering these complications, however, we focus on the problem of determining the opportunity cost of gas under assumed conditions of known (or predictable) supply and demand. The meaning of opportunity cost The opportunity cost for gas, or any other commodity, can be thought of as the price that will equate demand and supply. If the good is internationally traded, then the relevant import supply and export demand functions must be included in the calculations. An example of this situation is shown in Fig. 1. If the good were not traded, its demand and supply curves would be those labeled “domestic demand” and “domestic supply.” Its opportunity cost to the country would be P NTand the appropriate quantity to produce would be QNT. Once there is an international market in the good, however, the relevant demand and supply curves must take into account the import and export possibilities. If the good can be imported at a price P, and exported at P, (where the difference between P, and P, represents the freight, insurance and handling cost of trade), then the relevant demand and supply curves become the linked one labeled “traded supply” and “traded demand.” Their intersection is at the price P, where the quantity Q, will be produced, QC will be consumed and the difference (Qp - QC) will be exported. In this case, the availability of an international market means both that more should be produced and
DomeStIC SUPPlY
Tradable SUPPlY
Tradable Demand
I
Qc Fig. 1. Supply and demand
I
QNT for traded
I I
QP
c Quontiiy
and nontraded
goods.’
The economics
of natural
gas development
241
also that a higher price should be charged to domestic consumers than if there were no export market for the good. The net gain to the country from producing and exporting the amount Q, - Q, is greater than the net loss to the country of producing and consuming only QNT at the lower price. This simplified diagram demonstrates the importance of the tradablefnontradable distinction in determining opportunity costs. Natural gas, of course, is generally not a commodity that is directly traded by developing countries. However, the distinction is still relevant as long as the gas is used domestically to substitute for another commodity (such as fuel oil) that is tradable. Gas only becomes nontradable in the economic sense when, at the margin, additional supplies that could be produced can no longer find any local markets where they would be replacing traded goods. The demand curve
An example may help to clarify this point. Figure 2 shows a demand curve for gas that is derived from composite data from two studies on the unit value of gas in various possible uses for a middle-income developing country. The length of each “step” represents the net-back value of gas derived from the export or import prices of the goods it is used to produce. For this particular country the highest value uses of gas are for peak power generation (where it essentially replaces diesel oil), household distribution (where it replaces LPG and kerosene) and methanol. The total amount of gas that can be consumed for those purposes, however, represents less than 5% of the potential market even excluding LNG. The fuel oil and coal that can be replaced in the power and industrial sectors clearly constitute the bulk of the market. Production of fertilizer and steel based on gas is sized to replace all imports of those commodities and generate a surplus for potential exports until around the year 2000 when domestic demands are projected to be large enough to absorb the full output. The LNG net-back value is based on a 2 X lo6 ton/yr facility which was judged to be the maximum amount which the country could reasonably expect to sell. For these reasons, if one were to visualize how the curve in Fig. 2 would shift over time, certain steps would grow longer (i.e., those based on domestic demand such as household distribution and fuel oil replacement) while others (i.e. fertilizer, LNG) would probably remain unchanged. The supply curve
Turning now to the gas supply picture, Fig. 3 illustrates a stepped cost function where the length of each step represents an amount of sustained production that could be
A
I
1.oco Consumption
of Gas (MMCFD)
Fig. 2. 1990 gas demand.’
P. B~URCIER
242
costof Gas (SMMBTU)
Discovered
“Probable”
SUPPlY
SUPPIV
r-----0 l.MM
500 Supply
I.500
2.oca
of Gas (MMCFD)
Fig. 3. 1990 gas supply.’
delivered for the incremental cost plotted on the vertical axis. The first (lowest) step in this function shows an amount of 100 MMCFD (MMCFD = lo6 ft3/d) of onshore, associated gas available at an incremental cost of $0.20/106 Btu. The second and third steps show production of nonassociated gas from onshore and offshore fields, respectively, at progressively higher costs. The sum of these three steps gives the country’s projected 1990 gas supply based on today’s proven reserves. The dashed line, drawn as the fourth step, shows that the full cost of finding and producing an additional 100 X lo6 ft3/d from reserves presently classified as “probable” is estimated at $2.50/106 Btu. The supply curve shown in Fig. 3 is a simplified picture which abstracts from at least two important complications. First, it represents deliverable rather than potential supply, and therefore incorporates considerations of appropriate field depletion rates and possible infrastructure constraints. In actual practice, the steps of the curve would be less abrupt since it is usually possible to increase production somewhat through added compression or temporarily faster depletion, A second, and more important, qualification is that the costs shown in the curve do not include any component to represent the opportunity cost to the country of consuming its finite gas resources now rather than in the future. Estimating, and accounting for, this depletion premium or user cost is a complex but important task in countries where gas is likely to be in excess supply for a relatively short period.7 Having discussed the derivation of both demand and supply curves, we are now in a position to superimpose the two in order to determine the opportunity cost of gas to this country. Figure 4 shows the result. Based on potential production from proven reserves, 1990 gas availability would be sufficient to meet all of the uses down to and including steel production and about one-tenth of the coal substitution. This would indicate that the opportunity cost of gas would be derived from its value as a coal replacement, equivalent to about $2.00/106 Btu. At this value, the probable reserves would not be developed since the cost of that gas would be higher than the equivalent cost of coal while all of the higher valued uses for gas were already being served. If one visualizes these curves in 1995 or the year 2000, however, rightward shifts in the demand function would make it profitable to develop the probable reserves. In that case, the intersection t Work carried out in this area at the World Bank indicates that the depletion premium at the beginning a period of IO-15 yrs of supply surplus may still amount to 50-607’0 of the fuel oil equivalent price.
of
The economics of natural gas development
243
Price of Gas (S/MMBTU)
Discovered
“Probable”
SUPPlY
SUPPlY
a-
I I
-------
r-“-
o
p -----__
_____
j
L
-,J
1 A
-m-J
I
I
500
1,000 Quantity
I I.500
1 2.000
of Gas (MMCFD)
Fig. 4. 1990 opportunity cost of gas.’
of the supply and demand curves would be on a horizontal segment of the supply curve. This means that the opportunity cost of gas would be equivalent to its incremental supply cost (abstracting from the depletion premium issue mentioned above), and no coal substitution would take place. Problems with opportunity cost pricing While the economic advantages of opportunity cost pricing are clear and, as shown above, it is not an impossible empirical task to derive such prices, it must be admitted that few gas-consuming countries follow this approach. There are at least two reasons given which are examined below. First, pricing at the marginal opportunity cost means that intramarginal producers and consumers may reap large excess profits. For example, in Fig. 4 if gas is priced at $2.00/106 Btu, then the least-cost producer would earn (before tax) rents of $1.80/106 Btu. However, a profits tax is an efficient and practical device to ensure that these surpluses are returned to the rest of the economy. A uniform price combined with profits taxation simply parallels the universal practice for crude oil. (Further details of practical profits taxes are described in Ref. 4.) A second objection to opportunity cost pricing in some countries is that the prices of competing fuels may be subsidized (or, more rarely, taxed) at levels which would encourage uneconomic fuel choices. For example, if fuel oil were subsidized and sold at a price equivalent to $1 SO/ 1O6 Btu, then pricing gas at its opportunity cost of $2.0011 O6 Btu in the case shown in Fig. 4 would discourage fuel oil users from shifting to gas. Clearly, the best solution to this problem would be to remove the fuel oil subsidy but, if that is not immediately possible, the government may consider delinking the producer and consumer prices of gas in order to permit the consumer price to be competetive with that of fuel oil while retaining the producer’s incentive to explore and produce gas. IV.
NATURAL
GAS
EXPLORATION
INCENTIVES
Most exploration/production agreements do not include specific gas pricing provisions. Instead, there is often only a general provision that, in the event of a significant gas discovery, the investor and the government will negotiate a price. In many cases, the price negotiated under such agreements has been determined on a cost plus rate of return basis at a level adequate to justify incremental investments but insufficient to justify the
244
P. BOURCIER
initial high risk exploration. Given this experience, in many countries private oil companies have been unwilling to explore in gas-prone basins despite attractive geological potential. Providing a more explicit gas pricing provision which links the price to its market value is a necessary condition to induce greater exploration in these areas. The central problem in specifying a specific gas price prior to exploration is supply uncertainty. In advance of exploration, it is not possible to predict the size of future discoveries. If there is a discovery whose supply potential is large relative to domestic demand (but too small or costly for export) then the marginal value of gas in that country will be altered dramatically. This is illustrated in Fig. 5. Initially we assume domestic gas demand exceeds supply [Fig. 5(i)] with domestic gas production OQ, up to the point where marginal costs equal the price of the alternative input OP, [Fig. 5(i)]. We assume the supply-demand gap Q2Q1 is met by fuel oil. In Fig. 5(ii) we assume a significant gas discovery with costs of production below OP, which shifts the gas supply curve to the right to ABCE. Now, domestic gas supply completely displaces fuel oil, and the marginal value of gas is reduced to OP2 with production OQ . The diagram illustrates that whenever uncertainty about the supply curve of a nontraded good is large relative to domestic demand, then the expected price of that good is also very uncertain. Herein lies the difficulty of specifying explicit gas pricing provisions in exploration contracts. At the very least, however, such pricing provisions should establish the principles and methodology to be adopted in the event of a commercial gas discovery. Contractual language might specify only that the parties will agree, following discovery, on a price determined by reference to the marginal opportunity cost of the gas at the time. Although such a general provision is subject to interpretation, taken together with provisions for independent technical input and arbitration, this language has the merit of establishing the principle of opportunity cost pricing rather than cost plus pricing and hence altering investor expectations of the likely future price. It represents a considerable improvement over contracts which simply state the parties will negotiate a price without establishing any clear pricing guidelines, particularly in countries where there is a recent history of very low producer prices (e.g. Egypt and Pakistan). Another important step to promote exploration for gas in developing countries is to ensure contractors a right to export gas in excess of a specified national reserve, as was recently done in Egypt, and to receive the international price for those exports. This provides a clear, prior guide to the company that should it make a certain size discovery it will have access to those reserves for export. Although this type of national reserve provision needs to deal with problems of prorationing of export entitlements and independent audit of proven reserves, the case of Egypt indicates that solutions can be found. The main problem here is helping the government to determine the minimum necessary size of, or formula for determining, such a national reserve. Supply constrained countries
Developing countries can usefully be categorized into those which are gas supply constrained and those which are gas demand constrained. In the former case, a scarcity of supply relative to demand means that a predictable value can be placed on additional discoveries of gas on a delivered basis up to the limit of excess demand. In demandconstrained countries, domestic demand is small relative to proven gas reserves so the marginal value of gas in the ground approaches zero. Different approaches will clearly be required in the two types of countries to promote appropriate levels of exploration investment in each. In supply-constrained countries, guaranteed purchase provisions and gas buy-back provisions are two of the specific measures that have been used to promote additional exploration. The formula can be used whenever the government can predict with reasonable confidence the magnitude of the supply-demand gap in the medium term. Guaranteed price provisions in exploration agreements would provide a specific gas price formula for the next x million cubic feet per day of gas delivered where the quantity would be equal to the expected gas shortfall. For example, if a gas supply shortfall over the next decade in a country is expected to result in the conversion of a thermal power
The economics of natural gas development
245
m
8
‘C
a
246
P.
BOURCIER
generating plant from gas to fuel oil (or new plants are to be oil fired) then the gas price formula for a quantity of gas up to x X lo6 ft3/d would be linked to the price of fuel oil.? This pricing incentive (by transferring price risk to the government) would act as an important stimulus to exploration investment in gas-prone areas. The guaranteed price provisions would be combined with the profit sharing provisions applicable to oil to ensure that, for low cost, intramarginal producers, a large share of producers’ surplus was returned to the country. One problem with guaranteed price provisions is that they presume that newly discovered gas in a supply constrained situation should be developed. This is not necessarily the case where supply is constrained not by physical delivery limitations, but by the desire to maintain a certain production/reserves ratio. Often a new discovery will not be the least cost means of adding to supply; rather the discovery should be added to reserves for future use and the rate of production from existing fields with established infrastructure increased. But here, clearly, the national and investor interests diverge: For the investor there is no point in exploring just to add to national reserves. One solution, in this situation, is a buy-back provision where, if the investor discovers proven reserves in excess of some minimum size (say 1 X lOI ft3), and if the investor and government do not agree on a development program, the government undertakes to reimburse the contractor’s exploration costs in return for relinquishing the discovery. A sliding scale premium might be paid for larger reserves. Although reimbursement of costs would not, in itself, be sufficient to induce gas exploration, taken together with the potential for oil it would enhance the expected value of an area. The buy-back provisions would have to be limited to the quantity of gas which could be expected to have value to the country in excess of the cost of buy-back. It would also be necessary to specify in advance the reimbursable exploration costs. These should be limited to a particular program of wells in order to avoid creating an overincentive, at the margin, to drill additional wells merely to obtain reimbursement of sunk costs. Additional exploratory or appraisal wells by the investor beyond the agreed program would not be reimbursable. There would also be a need for independent audit of reserves. This type of approach, which has recently been adopted in Egypt, could be applied in a number of other supplyconstrained countries. The assurance of buy-back up to the level of desired national reserves combined with the right to export reserves in excess of that level would be a significant exploration incentive. Demand constrained countries Neither of these approaches work in demand-constrained countries where the absence of a predictable gas use means an ex ante pricing basis cannot be established. In demandconstrained countries, where proven gas reserves exceed requirements for domestic use but are too small or too expensive for export projects, the economic value of gas in the ground approaches zero. In this case, exploration contracts can only establish the principle of opportunity cost pricing. It may be argued that in demand-constrained countries there is no need to encourage gas exploration, but this is not necessarily the case in countries with no proven reserves and negligible domestic demand. It may be that only following a discovery can potential demand be converted into effective demand. However, since users will not negotiate deals just in case a discovery is made, the lack of certainty regarding the m.arket for gas is a significant deterrent to gas exploration. The fact that a discovery in such a country has a low economic value can attract footloose gas using industries such as fertilizer or methanol plants. In these situations countries should encourage such gas development projects whenever the estimated netback value of the gas in use exceeds the marginal investment costs of field development and transmission (plus any depletion margin that may be imputed to reserves in the ground which in these countries is likely to be very small). However, to attract these industries there must be an agreement over price. Potential users naturally seek to negotiate a low into-plant price arguing that a significant incentive is required to set up t The guaranteed price need not be 100% of the expected economic value of incremental supply. It could be 70% or even 50%, so long as the price remained above the expected marginal cost of exploration and development of new reserves by a significant margin so that producer incentives were retained.
The economics of natural gas development
247
in a developing country and that expected profitability is marginal. There is always considerable market risk confronting export oriented users. A low gas price, by reducing the downside risk, is often an important factor in persuading an investor to set up in one country rather than another. While the government wants to attract the user industry, the establishment of a low price runs the risk of transferring the major share of the benefits of gas ownership to the user should the project prove highly profitable. One solution to this predicament is a risk-sharing contract under which the gas price is conditional on the ex post profitability of the gas-using industry. This type of price arrangement contains two parts: (i) a base price, combined with a take-or-pay sales agreement set just high enough to cover the cash flow needs to amortize incremental field development and transmission costs (i.e., operating costs and debt service payments and possibly a return on equity); and (ii) a supplementary unit payment based on the ex post profitability of the gas-using industry that escalates with increasing ex post discounted cash flow (DCF) return on investment. This arrangement ensures that the government obtains a high share of the ex post rent from gas production while minimizing user front-end risk (and accelerating cash flow), thereby attracting the investment in the first place. Thus this approach can make the government competitive in attracting footloose industries to utilize gas that would otherwise have no value, without negotiating away permanently the national benefits. V. FINANCING
LDC
GAS
DEVELOPMENT:
TWO
EXAMPLES
Thailand and Egypt are two countries where significant progress in gas development has been made recently. The type of studies, technical assistance and external financing that was required in the two countries differed considerably, but the approach of dealing with project difficulties by developing a better perspective on broader energy sector issues was successfully applied by both countries. In both, the role of the World Bank was to assist the government in identifying the key sector issues and in formulating a program of studies and investments to address them. Thailand When the Government of Thailand approached the Bank in 1976, the situation was at a classical stand-off. The company exploring in the Gulf of Thailand had made an important discovery and was willing to proceed with appraisal drilling but only after a contract defining the conditions under which the gas would be sold to Thailand was agreed. The government, which had no previous experience in gas, was faced with a series of difficult decisions about production volumes and possible uses of gas as well as prices and financing arrangements. While the two parties had been holding discussions for some time, no agreement was in sight. At the request of the government a Bank mission visited Thailand and, jointly, with the Ministry of Mines, prepared a plan for a program of studies to provide the government with the basic data required to make the necessary decisions. Consultants were hired to evaluate the discovery and monitor future drilling, to evaluate the market for gas under various price scenarios and to advise the government during contract negotiations. The various studies concluded that there was a market for gas primarily as a substitute for fuel oil in power generation and that gas could be brought to the market under conditions which would be attractive to Thailand while providing an adequate return to the oil company. This led to a contract which enabled the company to proceed with delineation drilling. At this point, although considerable manpower had been expended, no financial commitment had been made. Once enough reserves were proven, the company financed field development while the government mobilized the financing required for the pipeline. In addition to the Bank loan of $ 107 X 106, this included export credits and commercial loans in excess of $ 300 X 106. The project was completed on time and well within the cost estimates.
In the case of Egypt, the government identified early in 1977 the potential economic benefits of substituting natural gas for liquid fuels in the domestic energy market, thereby
248
P. BOURCIER
increasing net oil exports. This recognition followed completion of a gas utilization study promoted and financed by the Bank. So far, the Bank has financed four gas projects in Egypt in support of the country’s national gas development plan. These projects deal respectively with the recuperation of flared gas in the Gulf of Suez, the construction of a domestic gas distribution system in Cairo, continued gas exploration in the Western Desert, and development of the offshore gas field of Abu Qir. Each of these projects included finance for technical and economic studies to address sector issues ranging from updating estimates of gas reserves to devising a better price structure for petroleum products. While relatively few cofinanciers have been involved in these projects, so far, we expect this to change now that the feasibility of gas development in Egypt has been established and that significant steps have been taken by the government to interest private oil companies in exploration in gas-prone areas, as well as to attract increased commercial financing for development and transmission investments. In general, World Bank financing covers only a fraction of total project costs. It is applied flexibly to finance items of less interest to commercial bankers. In particular, because of the longer maturities of Bank loans, they are often used to finance basic infrastructure which, while contributing substantially to the projects’ economic viability, may not generate immediate financial returns. Through the loan documentation, certain guarantees, cross default clauses for example, can also be provided to private bankers which may increase their willingness to participate in project financing. Recently the Bank has introduced several new financing instruments specifically designed to facilitate private sector lending to developing countries by taking advantage of the lower perceived lending risks resulting from association with the World Bank Group. Despite these new initiatives, the financing of gas development remains difficult in many countries. In the poorest developing countries despite attractive projects, the noncreditworthiness of the country precludes commercial financing. The Bank’s limited resources for soft loan financing (IDA) may prevent good projects from proceeding in precisely those countries which most need them.5 CONCLUSION
There is no doubting the substantial potential for natural gas to meet a growing share of developing countries’ energy demands. Our experience to date suggests that the optimal pattern of investment in the gas sector and appropriate gas pricing policies will vary widely among countries depending on their particular demand characteristics and supply endowments. Sector-wide supply and demand analyses will be a prerequisite for determining the marginal opportunity cost of gas which should be the determinant of the gas price. The absence of explicit gas pricing provisions in exploration agreements is a deterrent to exploration in gas-prone basins. Contracts should, as a minimum, establish the principle that gas will be treated in a similar way to oil with a uniform producer price based on the marginal opportunity cost and gas profits subject to similar profit sharing fiscal arrangements. In supply-constrained countries, more specific contract provisions can be devised to encourage exploration. The financing of gas projects in creditworthy countries does not create insuperable problems once an appropriate gas investment program is defined and pricing policies introduced which permit and encourage investors and financiers to commit themselves to the investment program. The central role of the World Bank is to assist governments in devising and implementing such an investment program. REFERENCES I. World Bank, The Energy Transition in Developing Countries (1983). 2. William F. Sheldrick, “The Effect of Energy and Investment Costs on Total Fertilizer Production Costs,” paper presented at ISMA Meeting, London, p. 6 (1981). 3. World Bank Energy Department, Energy Department Paper No. 4, “Alternative Fuels for Use in Internal Combustion Engines” (198 1). 4. K. F. Palmer, “Mineral Taxation Policies in Developing Countries: An Application of Resource Rent Tax”, IMF Staff Papers, Vol. 27, No. 3 (1980). 5. A. W. Clausen, “Address to the Board of Governors of the World Bank” (I 982).