Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

This chapter begins with a general discussion of why an economic analysis is required and whether it should be called economic or financial. The costs and cash flows for a single isolated geothermal project are then introduced, firstly without discounting the cash flow, to illustrate methodology, and then incorporating it, as is essential for the economic comparison of projects. The various parameters used for project economic comparison are explained, and sensitivity analysis discussed. The role of a geothermal power station in a network is considered next, followed by the need for a more widely based energy analysis. The chapter closes with a discussion of steam sales contracts.

10.1 Introduction

The scales of geothermal energy extraction vary widely. At one extreme lies the single-family dwelling which makes use of a nearby spring without the need to drill at all, and at the other, the government organisation or multinational company with hundreds of MWe of generating capacity installed at a cost of hundreds of millions of dollars. At the lower end, the funds are simply drawn from the family bank account if they are available, but at the upper end, economic analysis is required, and the decision-making process is complicated.

The upper end of the scale may be the efforts of the government to increase the national supply of electricity by making use of its geothermal resources, thus avoiding further dependence on fuel imports. The fundamental economic problem facing all countries is how to allocate limited resources (labour, capital, natural resources and foreign exchange) to a variety of different uses in such a way as to maximise the net benefit to society. Turvey [1968] deals with the problem of valuing the electricity that might be produced to see if the investment of resources and funds is justified. For developing countries, in which some of the best geothermal resources are situated, funding is provided by the development banks, e.g. the World Bank and the Asian Development Bank. Also at this upper end of the scale lie government decisions to invest in geothermal resource use in response to global warming concerns. Such decisions will in part be politically motivated, but must be supported by economic analysis to some extent. It is not always necessary for scientists and engineers to be very close to the economic analysis at this level. In the early stages of considering the use of geothermal energy at a particular location, it may be that the exploration and feasibility studies necessary to support the economic analysis form a well-defined package of information which can be passed on without further involvement, but eventually one particular project will be decided upon for further investigation, and wider involvement is necessary.

A project may be defined in general terms as the investment of funds to construct a facility that will turn out goods, the sale of which generates a stream of profits sufficient to recover the initial investment and more. There might be several projects in contention, representing alternative ways in which the goods might be produced, and it must be decided which to invest in. Well-established methodologies exist to provide numbers by which the projects may be differentiated, but they are accounting calculations in which only money is recognised. Other benefits and disadvantages must be considered in some other way.

In the engineering industries, chemical engineers seem to have made use of the methodology ahead of other branches, perhaps because chemical plants are often stand-alone facilities reliant on specific markets for their output. Power stations for bulk electricity generation are different. In the commercial world in general, manufacturers can control the supply of their goods to maximise their profit; however, electricity supply is regarded by governments as essential for the social and economic well-being of their nation, and they seek to limit manipulation by private enterprise. One way of achieving this is by having the government own the electricity generators, which until the 1980s was the case in the UK (the Central Electricity Generating Board) and in New Zealand (the New Zealand Electricity Department), examples both large and small. Alternatively, governments can provide regulations to limit the profits from the sale of electricity to a level that suits their policies, and in this way, the revenue stream to the generating companies is controlled, directly or indirectly. The significance of this here is that the standard methodology produces a comparison of projects in terms of their level of profitability, so it must be modified if profitability is regulated. It turns out that the modification is in viewpoint rather than numerical procedure.

The numerical procedures themselves are very simple—the time value of money, which most of us understand intuitively via the concept of interest, is the only arithmetic process beyond simple addition and subtraction. However, difficulty arises because the focus of attention is cash flow, which is modified by taxation and the financial rules and customs of the particular country or company concerned. It is impossible to perform an economic analysis of a real project without incorporating these rules, for example, depreciation, but both the rules and the terminology are foreign to most engineers. Fortunately, they are not fundamental to understanding the methodology, so there is no necessity to include them here, and the project examples have been stripped down to basics to illustrate the effects of time on the interpretation of cash flows. Economic analysis without the full accounting rules still has value at project management level.

This is an appropriate stage at which to discuss terminology—are the issues of this chapter economic or financial? The assessment of engineering projects for profitability is traditionally called economic analysis—thus Allen [1991] uses the title “Economic evaluation of projects” in his book which is applied to the chemical engineering industry. The book by Marsh [1980] is entitled “Economics of Electric Utility Power Generation” and includes “financial simulation” as one of the methods of analysis available. Turvey and Anderson [1977] define the difference; thus,

“A ‘financial’ appraisal of a project is made either to determine its capacity to service debt and contribute to subsequent investment by the borrower, or to determine the return to the investor. An ‘economic’ appraisal or ‘cost-benefit analysis’ is aimed at determining whether or not the project is in the national interest”.

A definitive answer is unimportant here, and the term “economic analysis” will be used, although the subject matter would be included in what is defined above as financial appraisal. No attempt will be made to explain purely accounting matters such as ownership (shareholding), taxation and depreciation. Neither are the various means by which funds can be raised for geothermal projects discussed. The complexities of getting together many parties to fund a project are illustrated by Ogryzlo and Randle [2005] in their history of events leading up to the financing of a Nicaraguan project Scientific and engineering matters play a part in this, for example, in the form of an independent opinion (due diligence examination) on the likelihood of the project producing the steam required and generating electricity at the predicted cost, but such studies are well defined and, as mentioned earlier, form discrete packages of work.

A power generation project in isolation is examined first, assuming that its main components have been selected by the organisation after considering alternatives. The aim is to illustrate the general pattern of cash flow and how it can be affected by decisions and events that occur throughout the project life. Methods of quantifying project performance which are needed to decide between options are then addressed. The remainder of the chapter deals with related economic issues relevant to scientists and engineers.

10.2 A Single Isolated Project

A geothermal power project is a long-term undertaking involving the outlay of a large sum of money which generates no income over a period of several years during construction, followed by income from the sale of electricity as smaller sums over a much longer period. Figure 10.1 shows a spreadsheet with the first column representing the year of the project and the other columns representing, for every year, the cost incurred, revenue earned, the sum of these and the debt remaining, assuming all earnings are used to repay debt at the end of each year. In reality, the project plan, which is the list of items up to year 5, will be very much more detailed, and it is better to lay out the sheet with a column for each year, entering the cash items in rows.

Fig. 10.1
figure 1

A simple spreadsheet for a single project, lacking interest charges. The graph responds instantly to the change of any parameter so the sensitivity of the date of zero cumulative cash flow can be investigated

Full size image

Figure 10.1 is a simple example made up for a 20 MWe project. It is (deliberately) deficient in that the cost of capital is zero, and there is no interest on accrued earnings. The cost data were real estimates at 2007, taken from a report commissioned by the New Zealand Geothermal Association [2009]. Standard accounting practice for recording transactions—see, for example, Marsh [1980]—is not used; instead, income is shown positive and expenditure negative. The vertical rows represent time in years. The activities during the construction period of 5 years are itemised and costs assembled based on the following:

  • Gross output of station = 20 MWe

  • Cost of station at $2.2 million/MWe = $44 million

  • Net output is 94 % of gross = 18.8 MWe

  • Total wells required = 8 (5 production and 3 injection)

  • Cost of drilling = $5 million/well

  • Steamfield cost = $18 million

The operating and maintenance costs during the life of the station are assumed to be $2.2 million per annum.

In the first project year, the costs are for exploration and planning, assumed to be $3.5 million. To the year 3 drilling costs, $2 million has been added for preliminary design and permitting. The major capital items of steamfield and power station itself have been assumed to be paid for in two parts, 30 % in year 4 and the balance in year 5.

The station is assumed to operate at a capacity factor of 95 %, that is, it operates at full output for 95 % of every year. However, in the first year of operation, a pessimistic capacity factor of 50 % has been assumed to allow for testing and early operating problems. If the selling price of electricity is P c/kWh, the annual revenue is calculated as follows:

$$ \mathrm{ Annual}\ \mathrm{ revenue}=\mathrm{ net}\ \mathrm{ output}\ \mathrm{ rate}\ \left(\mathrm{ kW}\right)\times \mathrm{ capacity}\ \mathrm{ factor}\times 8760\ \left(\mathrm{ hrs}/\mathrm{ year}\right)\times \mathrm{ P}\left(\mathrm{ c}/\mathrm{ kWh}\right) $$
$$ =18.8\times 0.95\times 8760\times \mathrm{ P}/100\ \left(\mathrm{ dollars}\right) $$

The operating and maintenance costs have been deducted from the revenue for each year, and the remainder has been paid to reduce debt.

In reality, interest will be charged on loans. Funds will not simply be paid to sit in bank accounts until the cost arises; instead, detailed arrangements for making them available will be required, and yearly accounting is unlikely to be adequate. However, the purpose of this simulation is simply to show the items making up the cash flow. A graph of cumulative cash flow has been added, and using Excel©, it can be made to respond immediately to changes made to the chart. For example, the selling price of electricity can be entered into a single cell which is picked up in the calculation of revenue for every year. The spreadsheet can thus be used to see the effects of various changes, to either the parameters needed in the calculation which are linked to the value in a single cell under the heading of “data” on the sheet or changes to the cash flow items column. Changes worth examining, for example, might be:

  • A delay of 1 year after year 4, in which nothing is done

  • The need to drill one replacement well in year 10 and another in year 16

  • A decrease in the annual load factor from 95 % to 80 %

  • A decrease in the full load net output to say 15 MWe (as a result of steam supply shortages)

The cash flow pattern shown in the graph of Fig. 10.1 has the characteristic shape of all large projects; the debt builds up gradually at first and then steeply (even more steeply when interest is included), reaching a maximum (minimum on the graph) which coincides with the time at which income begins to be generated. Income from the following years reduces the debt gradually until it is paid off, when the cumulative cash flow curve crosses the zero axis. The time from first expenditure to this point is the payback period, and at this time, the project can be said to have broken even. After this, the net cash flow is positive and profit is being made for the remainder of the project life. Figure 10.1 coupled with Table 10.1 shows a break-even point 16 years after the first year of production if the electricity is sold at 8 c/kWh.

Having set up the programme, it can be used to investigate the economic effect of any changes to the project plan, but it must include the proper valuation of money, which is addressed next.

10.3 Economic Evaluation of Projects

All major power projects require significant starting investment over a construction period of a few years and might take 15 years or more to recover that investment. It might seem odd that the economic performance of such projects is still mainly assessed by reducing the considerable cash flows over the project life to a single parameter and then comparing the values of this parameter for all the projects being considered. According to Leung and Durning [1978], the economic comparison of projects in the electricity supply industry as a whole was rather empirical until the 1950s, after which the methods described here came into use. It must be remembered that the prediction of events and cash flows tens of years into the future is accompanied by uncertainties which even detailed analysis cannot eliminate. Economic decisions call for judgement and do not rely only on the result of a formal analysis. A committee responsible for selecting alternatives would equip itself with several of the single parameters but then exercise its collective wisdom.

10.3.1 Discounting the Cash Flow

The single parameter methods which involve the timing of costs and revenues can be described as discounted cash flow methods. Figure 10.1 shows the cost of the first item (exploration and planning) as $3.5 million. If interest was charged on this loan at 5 % pa, then without any further expenditure, the debt in successive years becomes $3.68 million after the first year and then 3.86, 4.05, 4.25 and so on. The value in any year is the initial expenditure times (1 + r)t where r is the interest rate on the debt and t is the number of years since the debt was incurred. Suppose that it was decided to put money aside now to spend on this work 3 years into the future, what sum would be sufficient? The answer is just enough to grow to exactly $3.5 million at 5 % pa, i.e. by an annual multiplying factor of 1.05; the sum is $3.02 million. This number is the discounted value of the expenditure or its present value, “present” meaning at the time the money was set aside. The concept is that of interest, but the terminology is different because sums of money are being brought backwards in time rather than forwards, and they shrink rather than grow. Thus the present value, PV, of a future amount Q is

$$ \mathrm{ PV}=\mathrm{ Q}/{\left(1+\mathrm{ r}\right)}^{\mathrm{ t}} $$

The factor 1/(1 + r)t is called the discount factor, and every cost or revenue throughout the project can be brought backwards in time to a single figure valued at year 0 (or any other year) by multiplying it by the appropriate discount factor. What was called the interest rate in working forward in time is called the discount rate in working backwards, and they may not have the same value. Money is not necessarily borrowed from a bank, it might be raised as bonds by a company who has many uses for it apart from building a power station and its value is represented by the rate at which profit could be made if it was used for other things. This might be higher than a bank lending rate. The discount rate is often called the “cost of capital”.

10.3.2 Net Present Value

A new spreadsheet incorporating discounting is shown as Fig. 10.2. The discount factor for each year is shown in column H, calculated from

Fig. 10.2
figure 2

The single project of Fig. 10.1 incorporating interest and discount factors

Full size image
$$ \mathrm{ Discount}\ \mathrm{ factor}=1/{\left(1+\mathrm{ i}\right)}^{\mathrm{ t}} $$

where i is the discount rate and t the year number, given in column A. The formula in the spreadsheet for year 8 is

$$ \mathrm{ Discount}\;\mathrm{ factor}=1/{\left(1+0.01^*\$\mathrm{ C}\$33\right)}^{\wedge}\mathrm{ A}8 $$

so a change in discount rate stated as a % in cell C33 transfers through the whole spreadsheet.

The annual net cash flow is the sum of costs and revenues. The costs and revenues are as defined for Fig. 10.1 but are entered in this table as end of year figures, without any interest. All transactions are assumed to occur at the end of each year, and no interest is incurred during that year. The annual net cash flow is the sum of costs and revenues, and by applying the discount factor for that year, a column of NPV values for every year is obtained—these are the values at the present time (year 0) of the net cash flow in any future year. These values are accumulated in column J, from which it can be seen that all costs have been recovered by year 20 with a cost of capital of 10 % and an electricity selling price of 12.5c. This would probably not be an attractive proposition, and a higher selling price and lower cost of capital would produce a higher cumulative NPV, which is a $ value representation of the project, one idea of the profit.

If the selling price of electricity is regulated, directly or indirectly, the cumulative NPV at the end of the project is not very helpful for choosing between alternatives—all NPV’s are likely to be similar. Comparing NPV’s would be helpful for the producer of unregulated price goods, because the option with the biggest profit would be most attractive. For the electricity industry, a method known as the “present worth of revenues” was introduced in the USA (Marsh [1980]); it represents a change in viewpoint rather than a significant change in calculation procedures. An acceptable return on investment is chosen and applied to all of the options being considered, and annual payment of this return becomes a cost to the project. Instead of the net annual income being the difference between revenues calculated for the expected electricity price and costs, the net annual income is decided upon first and added to the costs, which leaves a figure for the revenues to be collected from the sale of electricity. The option which has the lowest required revenues is the preferred one.

10.3.3 Payback Period

The simplest of the single parameters by which to judge the economic performance is payback period, this being the time after the start of expenditure at which the project has earned enough revenue to pay back the money invested—it was defined with reference to Fig. 10.1, although its calculation there did not involve discounting. For the conditions shown in Fig. 10.2, the payback period is 20 years.

In comparing options, the alternative with the shortest payback period is supposedly preferred. Making payback period the sole criterion would effectively make cancellation of the debt the absolute priority. There could be legitimate reasons for this, perhaps a need to have the capital available for another project as soon as possible or because the project is risky and the shortest possible exposure to risk is called for. The revenues earned by the project up to this time are included in the calculation, but later revenues are not, so using payback period as the single criterion makes no attempt to estimate the benefits of the project as a whole.

10.3.4 Internal Rate of Return

There is a discount rate at which the project of Fig. 10.2 only just breaks even if the life is 20 years and the cost of electricity is fixed at 14 c/kWh. This can be found by adjusting cell 31 and watching the final cumulative NPV, and it turns out to be about 12 %. This discount rate is called the internal rate of return (IRR) or the discounted cash flow rate of return. It has been calculated here from the net annual cash flow stream, but it can also be explained as the discount rate at which the present value of the cash flow stream of costs exactly equals the present value of the revenue stream over the project life. The order of the arithmetic operations, finding the net and then discounting, makes no difference. It marks the discount rate at which the project changes from one that makes an economic profit to one that makes an economic loss. The project might still be justifiable by providing non-economic benefits.

The IRR is informative but limited in the information it provides. The project life must be defined for this calculation, from which it follows that the IRR is an improvement over payback period as a representation of the project. If the cost of capital were to reach the IRR, the project would only just break even. The more the cost of capital falls below the IRR, the greater will be the NPV at the end of the project, so as great a margin as possible would normally be sought. However, as stated, the cost of electricity is usually regulated. For electricity generation, it might be said that an IRR value well above the cost of capital would provide, not a high profit but a high margin of security against the risk of unexpected problems and hence costs.

In summary then, the IRR is used for comparison with the cost of capital; its calculation does not consider net revenues so cannot arrive at a $ valuation of the project.

10.3.5 Levelised Cost of Electricity

It was demonstrated above that by adjusting the discount rate by trial and error, the discount rate at which the project would just break even could be found. In the same way, the cost of electricity could be varied to find the value of discount rate at which the project just breaks even. This value is called the levelised cost of electricity, and it is defined as the sum, over the project life, of discounted costs divided by discounted sales in kWh:

$$ \mathrm{ Levelised}\ \mathrm{ cost}\ \mathrm{ of}\ \mathrm{ electricity}=\frac{{\displaystyle {\sum}_{i=1}^n\frac{{\left( \cos ts\right)}_i}{{\left(1+r\right)}^i}}}{{\displaystyle {\sum}_{i=1}^n\frac{(sales)_i}{{\left(1+r\right)}^i}}} $$

The levelised cost of electricity is commonly used to compare alternative methods of generation; like IRR, it provides some guidance as to possible profitability. The lower the levelised cost compared to the price at which the electricity can actually be sold, the greater the profit margin.

10.4 Factors Affecting Project Life

Project life is important in calculating project NPV and IRR. The life of a geothermal project depends primarily on the life of the geothermal resource, equivalent to the fuel for a fossil-fuelled station. Unlike fossil fuels, geothermal energy is not transportable over long distances, and the station must be built near to the resource. If the supply runs out altogether, there is usually no alternative but to sell the power station equipment. More likely, the supply of geothermal fluid or its temperature will decrease over the life of the project. The Wairakei station, NZ, was constructed with three stages of turbine, high-, intermediate- and low-pressure machines. Several years after commissioning, the delivery pressure of the steam from the wells decreased so the high-pressure turbines were removed and used at another resource.

Heavy mechanical engineering equipment is often regarded as having a 25-year life, although this is a rule of thumb rather than an accurate determination. The critical factors are wear, corrosion and mechanical stress (fatigue), and the level of detailed calculation of these factors that has gone into the design depends on the particular industry. In today’s aircraft industry, weight is critically important from the point of view of performance, and if structural weight is to be minimised, the life of components at given stress levels has to be exceptionally predictable to ensure passenger safety. In comparison, the Wairakei power station turbines were commissioned in 1956 and are still operating in 2012 after 56 years, because the manufacturer had to allow a margin to ensure the performance and safety of his equipment, and it was not possible to calculate this margin accurately at that time.

10.5 Other Economic Aspects: Sensitivity Analysis and Risk

The suggested use of a spreadsheet to examine variations in the proposed project is in fact a sensitivity analysis, in that it can be made to show the effect on the cash flow curve of any changes from the planned project or variations in the cost of services and items. Sensitivity analyses also need to be carried out at the project selection stage when single parameter representations of the project are being considered. There are several modern methods of illustrating risk—and hence sensitivity—visually, such as “tornado diagrams” and “spider plots”, but the basic aspects are as covered here.

As an example, consider the sensitivity of the cumulative NPV at year 20 to variations in the cost of drilling a well and the annual load factor of the completed station. The spreadsheet of Fig. 10.2 allows the cost per well to be called from a single cell (C34) whenever it is required to produce an annual cost. The cost has been varied in increments of 10 %, both positive and negative, applied to the datum case value of $5 million per well, and for each value, the NPV has been recorded. The results are shown in Fig. 10.3.

Fig. 10.3
figure 3

Sensitivity of NPV to variations in drilling costs and load factor over the last 5 years, for the data of Fig. 10.2

Full size image

The cumulative NPV has a value of $10.46 million after 20 years in the datum case of Fig. 10.2 (14.0 c/kWh selling price and 10 % cost of capital). It decreases as well cost increases and vice versa, and the variation in NPV is linear. The graph also shows a test of the sensitivity of NPV to load factor, for example, due to reduced resource output. This has been simulated by reducing the load factor over the last 5 years; Fig. 10.3 shows that a severe load factor reduction of 40 % in the 95 % value taken as the datum case over this relatively short period would reduce the NPV to almost zero, at which the project would only just break even—in purely financial terms, it removes the profit from the entire project.

Sensitivity analysis is discussed in more detail by Allen [1991]. More recently, Sanyal [2005] considered the sensitivity of geothermal electricity generation costs to various factors and also addressed the minimisation of levelised generation costs using enhanced geothermal systems (Sanyal [2010]).

10.6 Consideration of a Geothermal Station in a Network

A network is the group of interconnected power stations available to an organisation to supply electricity. The demand for electricity, referred to as the electrical load, is likely to follow a regular pattern which must be determined, particularly in bringing electricity into a new rural area. In Western countries, the domestic electrical load typically increases very rapidly from a low level overnight to a morning peak as users begin their day. The load decreases after breakfast, increases to a high level for a short period at lunchtime then decreases in the afternoon, peaks at evening mealtime and returns to its low level overnight. Industrial loads will be virtually continuous for some activities such as mining and ore processing and variable for commerce and light industry, peaking between breakfast and lunch and then again in the afternoon until work and business ceases for the day. The precise load at any time is unpredictable, being subject to random influences. The single electricity generating company considered here might have several power stations, some fossil fuelled (coal, gas or diesel), one geothermal and one hydro-station. It is not possible to store electricity per se, so the total output of the stations in operation must be controlled to exactly match the total load. Electricity can be converted to some other form of energy that can be stored—the spinning turbines of all the power stations in a network provide some storage like this, and flywheels have been proposed but are not used. Pumped storage is an option, in which excess electricity drives pumps to move water from a low-level lake to a high-level lake so that it can be reconverted to electricity by running the water back down again through a water turbine. From a purely engineering point of view, this is attractive because the pumping of water and generation of electricity are both mechanically efficient processes, so not much energy is lost. Few have been built presumably because their capital cost is unjustifiable.

The total load may be unsteady, but a large proportion of it may be constant. A load–duration curve is produced for the system, which shows how many MWe of the total load is always present, and this is referred to as the base load. The available power stations are ranked according to the cost of electricity produced and their ability to respond quickly to load changes, so each is appropriate to some part of the load–duration curve.

10.6.1 Marginal Cost and Load Following

The faster a car goes, the greater the fuel consumption per kilometre travelled; in economic terms, the speed increase has an associated cost increase which is called a marginal cost increase. When several stations are available to supply an increase in load, the one with the least marginal cost should be used first. This is not purely an economic factor, however, but depends on the rate at which any of the available stations can increase its load—axial steam turbines are limited in the rate of rise of temperature that can be applied without causing mechanical problems due to thermal expansion.

10.6.2 Base Load Service

In general, power stations are an integral part of an electrical transmission network distributing electricity to various loads. Such a system may be owned and operated by the government of the country or by commercial companies.

Figure 10.4 shows the difference in the pattern of capital and operating costs between a fossil-fuelled station and a geothermal station.

Fig. 10.4
figure 4

Comparative cost patterns for fossil and geothermal power stations

Full size image

The fossil-fuelled station was built over a period of 3 years and requires a supply of fuel throughout its life, which has been assumed to be 25 years. There is a significant cash flow throughout the station life due to the purchase of fuel. The geothermal station was built at a higher cost over a period of 6 years; this includes the exploration and drilling to confirm that the geothermal resource can supply enough geothermal fluid to power the station fully over the 25-year period. The geothermal station is “capital intensive” in that the bulk of the 25-year costs is invested before the station begins its output. The cost has been incurred whether the station is used or not, whereas the fossil-fuelled station avoids the fuel cost if it is not used. Geothermal stations should be operated continuously, to provide for that part of the load that is always present, which is referred to as base load. Geothermal stations have this capital intensiveness in common with nuclear power stations, which are also used to meet the base load, and for different reasons, neither type is capable of the fastest load following performance, which further favours their base load use. Hydro-stations are also capital intensive and, in principle, have zero fuel cost, like geothermal, so would-be candidates for meeting base load. However they are capable of faster load following, and their fuel supply is seasonal and unpredictable from year to year so must be conserved; both of these factors count against their automatic use to meet base load.

10.7 Energy Analysis

In New Zealand at the present time, it is possible to buy small cast iron water pipe fittings at a cost which would not pay for the electricity to melt the metal they are made of, let alone its machining and transportation to market. Viewed in isolation, the selling price is uneconomic. The reason may be that having set up the foundry and embarked on castings requiring a large amount of metal to be processed, the marginal cost of producing small items is low enough to make it profitable. Large projects require the same scrutiny as to value for money, but an assessment is not possible intuitively. This issue arose in the 1970s in connection with the UK nuclear electricity generating programme, when it was raised by Chapman [1974]. At the time, oil supplies were restricted worldwide, and it was proposed to build nuclear stations to provide alternative sources of electricity. This required the construction of a series of nuclear reactors fuelled by refined uranium and moderated by heavy water (D2O), both the product of energy intensive processing and thus requiring a supply of electricity from the existing fossil-fuelled power stations. Chapman’s message (see also the response by Wright and Syrett [1975]) was that although one of these reactors could produce more energy than was used in its construction, building many of them at a high rate did not produce the required result of decreasing the national dependence on fossil fuels. The programme, it was argued, absorbed all of the generated output and more before it was completed, which would have been many years into the future.

This issue is better understood today, for example, it has recently been examined by Frick et al. [2010] in an analysis which they refer to as “life cycle assessment”. Their reasoning parallels that of Chapman, namely, that political efforts to reduce greenhouse gas emissions and the consumption of finite energy resources have led to proposals to extract low-temperature geothermal water for electricity generation via binary power plants. The conversion efficiency is low, governed by the second law of thermodynamics, while high-grade (high-temperature) energy is required for well drilling, pipelines and manufacture of the power plant. The auxiliary power consumption of binary cycle plant is relatively high, and low-temperature geothermal fluid sources require fluid to be produced by pumping. All of these factors extend the project payback period. Inevitably, since the problem is so complex by involving, in principle, every energy consuming activity which makes a contribution to the power station construction, the analysis by Frick et al. draws only general conclusions. The value of their contribution lies in continuing to draw attention to the issue and demonstrating how to examine it.

10.8 The Steam Sales Contract

There are a few geothermal power projects in which the wells and pipelines are owned by one party and the power station and transmission lines by another. This arrangement splits an otherwise seamless engineering undertaking into two separate parts which are required to cooperate very closely. The split is not physical but organisational – Fig. 10.5. Each party has its own staff and management structure, and their interaction is defined in the Steam Sales Contract. Several physical points are defined at which the responsibility changes from one party to the other, for example, a point on the main steam pipeline and a point on the pipeline returning condensed steam to the field for disposal. It is convenient to consider that the power station has a “fence” around it, in terms of the boundary of responsibilities.

Fig. 10.5
figure 5

Organisation chart for parties served by a Steam Sales contract

Full size image

Referring to the parties as the steam supplier (Resource Company) and the station operator (Power Station Company), the purpose of the contract is to arrange for a reliable supply of steam sufficient to run the power station at full capacity. Assuming base load operation is intended, the contract will state the intended output in MWe and the steam conditions and flow rate required to produce this. The station operator pays the steam supplier for the steam actually supplied at a defined quality, so measurement equipment must be installed at an agreed point near the (hypothetical) fence. The measured parameters are likely to be flow rate, pressure and dryness fraction, assuming the steam to be saturated, and the flow rate and chemical species of any liquid carried with it. The steam will include gas, which generates power as it passes through the turbine but requires power to remove it from the condenser. If the gas flow rate turns out to be variable in the long term, then the contract may have to be written to account for it. The measurements and all related data, instrument specifications, etc. must be available to both parties, and the water properties (steam tables) to be used for calculations related to the contract should be specified together with the calculation procedure. The instrumentation must be maintained to keep the measurements within the manufacturer’s tolerance. The power station output will need to be measured at a defined point because the voltage is usually stepped up in stages downstream of the alternator, and each transformer introduces losses; the input terminals of the first transformer after the alternator might be chosen. Since the purpose of the measurements is to define how much the station operator pays to the steam supplier, a target amount per month or quarter is likely to be specified, with variations in payment if the actual amount supplied differs from the target. Both parties require an incentive to keep their performance to the contracted expectations, and any shortfall should therefore result in a penalty. Output in excess of the target might be attractive to the station operator, depending on arrangements for the sale of the electricity, in which case the payment formulae will include this possibility.

The arrangements at the end of the contract period must be defined. The contract may be renewable, but if not, then ownership of the assets constructed by the Resource Company must be addressed in the contract.

There are secondary matters to be contracted for in addition to the supply of steam. Electricity is required to operate the wells, pipelines, instrumentation, water supply and disposal pumps and staff facilities, and since geothermal stations are often at remote locations, the power station usually provides for this load. The load must be defined in the contract since the value of the electricity supplied must be accounted for in the transaction.

Depending on the location of the fence and the arrangements for separating water from the steam supply, separated water might have to be pumped out to the field from the station. Steam condensate will certainly have to be disposed off from within the fence.

Increases in the cost of drilling, casing and labour, for example, required for the execution of the contract, can be accommodated for by reference to national economic indices, to avoid the contractor building in contingency arbitrarily.

The details of the contract just described are relatively straightforward to define for an existing reservoir development and power station, where the resource is understood and the plant already in existence. Sometimes a steam sales contract is foreseen from the outset, perhaps due to an inability of one of the parties to raise enough capital for the entire project. Design aspects of the facilities will then need to be discussed. For example, steam jet ejectors use steam but have a lower capital cost than mechanical gas extractors, which consume electricity. There is an incentive for the Power Station Company to reduce its capital outlay and its parasitic electrical load by using ejectors, but this would increase the steam demand, to the detriment of the Resource Company. The parties involved would reach agreement on such issues and on the form of the contract before any related expenditure is committed to. The length of time for capital recovery is long, and each party needs to be certain that the other can perform as required.

Few steam sales contracts are published, but Puente and Andaluz [2001] reviewed the performance of a steam sales contract at the Cerro Prieto resource between the Mexican public electricity company, CFE, and a private contractor. The contract was for a period of 10 years beginning in 1991 and was for the supply of 800 tonnes/h of steam; it appears that CFE already had its own wells and steam supply system for the power station. It was awarded following a call for tenders and functioned well for several years until Mexico suffered an economic crisis as a result of which the contractor’s supply fell below the contracted rate, incurring financial penalties which made its position worse. CFE’s own steam supply declined due to lack of investment, causing CFE and the contractor to reach agreement with a second (US) contractor in the form of a steam supply contract for 1,600 tonnes/h, twice the original contracted flow rate. At the time of writing of the paper, the contract term had expired, and a new contract was being sought.