Electricity generation costs

5.2.1 Overall Goal and Purpose

The overall goal of PSI’s update to the previous report on costs and potentials for SFOE Energy Perspectives (BFE Energieperspektiven; (Hirschberg et al. 2005)) is to provide a broad, objective and impartial analysis of the relative characteristics, potential, and cost of the full range of future electricity generation technologies. The purpose of the economic analysis effort within the PSI update is to analyze the internal generation costs of the different technologies, based upon the trajectory of costs and generation over the lifetime of each unit. The economic analysis also has the goal of analyzing each technology with a common methodology, and using a common framework of shared data assumptions. As far as possible, the economic analysis also has the goal of applying a consistent level of moderate optimism to expected technological learning and advances based on the current maturity of technologies.

5.2.2 Procedure

5.2.2.1 Levelized generation costs

The levelized cost methodology (also called “Life Cycle Costing”, LCC) uses financial discounting to bring all construction costs forward, and all future costs backward, to the date of the plant’s start of operation (Figure 5.1). A uniform discount rate of 5% has been used for quantification of LCOE of all technologies.¹²² Future costs include operating costs (fuel, and fixed and variable operation and maintenance costs), as well as costs for plant dismantling, site restoration and waste treatment and storage costs. The net present value is then amortized over the generation lifetime of the plant, as shown in the Figure 5.1 below.

The annualized cost is then divided by the expected annual generation, based on an expected capacity factor or dispatch simulation.

122 Evaluation of LCOE of biomass conversion technologies includes external case studies for comparison; these case studies partially use different discount rates, which are explicitly mentioned.

187

Figure 5.1: Scheme of LCC methodology, resulting in average generation costs per kWh of electricity, equal to “Levelized Cost of Electricity - LCOE”.

Both current and future LCOE are quantified. Current LCOE refer to hypothetical new power generation units “to be built today” purchasing new technology on the market. In case of large hydropower and nuclear power, also the generation costs of the currently existing power plants in Switzerland are provided, since these plants contribute the vast majority to the current domestic generation will be part of the Swiss generation mix for many more years. In addition, “to be built today” is a theoretical concept for nuclear and (to some extent) large hydropower, since new power plants require extensive planning and licensing, are subject to complex approval procedures, and construction periods are substantial.

5.2.2.2 Cogeneration

In cases where electricity is produced by cogeneration (that is, both electricity and heat are produced as co-products), it is assumed that some of the heat can be sold, or that the heat can replace heat or fuel that would otherwise be purchased. In either case an appropriate heat credit is applied. This heat credit will generally be larger for avoided costs (e.g., natural gas fuel or district heating not purchased) than it will be for heat sales (heat is sold back to a district heating network for a much lower credit than the district heat purchase price). The heat credit is based on the expected annual heat revenue or avoided cost, as there may be no need or market for heat produced during some months.

For example, a geothermal plant with an electric capacity of 5 MW will produce about 35 MW of heat at about 14% electric efficiency. Annual heating demand has a load factor of about 20%, which means that about 60 GWh of heat could be used or sold annually. If the geothermal plant can replace 40 GWh of heat from a district heating loop at about 70 CHF/MWh (by direct self-use or direct sale to a nearby customer), and sell the remaining 20 GWh back to the district heating system at 10 CHF/MWh, then the annual heat credit will be about 2.8 MCHF from the savings and 0.2 MCHF from the sell-back. Obviously, depending upon how much of the heat can be used onsite or sold nearby, the heat credit could vary

188

widely from 0.6-4.2 MCHF. Likewise, if more of this low temperature heat could be used per year to displace district heat (e.g. an 80% load factor preheating material in a cement kiln or another industrial process), then the value could be about 240 GWh at 70 CHF/MWh, or almost 17 MCHF.

In general, LCOE both with and without heat credits are provided. Whether these are likely to be credited will in reality mainly depend on generation plant locations and amounts of heat generated; also economic incentives and regulations could play a role. Technology specific factors in this respect are discussed in the technology chapters and results are discussed from this perspective as well.

5.2.2.3 Applicability and potential limitations

By using the average cost per kWh, the levelized cost method allows comparison between small and large plants, as well as comparison between inexpensive plants with relatively high fuel costs and low capacity factor (e.g., gas turbines) and expensive plants with cheap fuel and high capacity factors (e.g., nuclear). The method does omit site and jurisdiction specific factors, including taxes, depreciation, insurance, regulatory costs, etc., but it does provide a commonly accepted benchmark for cost comparison. The focus of this method is on costs rather than revenue, but if different technologies produce electricity sold at different prices this can also be taken into account. Sensitivity analysis is used to show the dependence of the average cost upon variation in the difference cost components, as well as the plant life, interest rate, capacity factor and other variables. Figure 5.2 shows an example of sensitivity analysis performed for photovoltaics.

Figure 5.2 Example of sensitivity analysis: LCOE of photovoltaics.

189

Within PSI’s framework for energy systems analysis, the average cost of generation is used as the chief economic indicator of internal system costs, but other economic indicators (e.g.

total cost related to financing risks, or the fuel cost fraction related to energy dependence risk) are also calculated. The economic indicators are combined with environmental and social indicators for overall sustainability analysis. This includes tradeoff analysis between competing indicators, total costs based on monetization of environmental burdens (external costs) and multi-criteria decision analysis.

5.2.3 Estimation of future development of fuel prices

“It’s difficult to make predictions, particularly about the future” (variously attributed to Yogi Berra, Niels Bohr, Mark Twain, Samuel Goldwyn, and many others).

In order to place the economic evaluation of different generation technologies on a consistent basis, it is necessary to have appropriate and consistent fuel price assumptions for the major fuels used for thermal generation, in particular natural gas, coal, (oil) and nuclear. Biomass fuels (wood and non-woody) are dominated by more local pricing, due to lower energy density and lower economic transportation radius and are not further discussed in this section.

This section surveys the available future fossil and nuclear fuel price scenarios, describes their comparative results and chooses future price trajectories for use by the individual technology analyses and in the unified comparison presented in this chapter. It is important to realise that future price projections represent the best efforts of each projecting agency, but they are not firm predictions. Likewise, in this survey PSI is choosing price trajectories that are a scenario for future prices, recognising their uncertainty without estimating any associated probability.

5.2.3.1 Requirements

The major characteristics desired in the fuel price scenarios surveyed included the following:

• Long term. For long-term generation sector planning based on plant lifetimes, demand growth, technology changes, etc., it is necessary that the fuel price scenarios should also be quite long-term (out to 2040 or beyond). The majority of price scenarios are dominated by the short and mid-term time horizons due to commercial needs, which can also lead to shifts in definition of terms (some sources consider 5 years to be long-term).

• Multi-fuel. For consistency, it is by far the best if price projections are taken from a single, coordinated modeling effort, reflecting consistent assumptions for demand, resources, demand, substitution, etc.

• Regional differences. Fossil fuels have regional price differences (oil least, and gas most), based on quality differences and transport costs. Nuclear fuel is more energy dense, and so any regional cost differences are smaller, and more related to markets than fuel quality. These regional differences would again be accounted for in a proper integrated model environment.

These requirements imply that a long-term, multi-region or global, multi-fuel energy sector model will be required. Long-term demand growth is driven by assumptions about population growth, economic growth, energy intensity and price elasticity. Energy supply is driven by known and estimated resources, their costs of production, and the costs and

190

efficiencies to refine, transport and transform energy into services. Models find the prices at which supply meets demand, taking into account fuel and technology substitutions, price feedbacks, investment constraints, and other factors. Such models are complex, data intensive and require expertise, such that they are likely to be performed (or at least funded) by either a major country or an international agency.

5.2.3.1.1 International organizations/energy models

Two main, major energy models providing long-term price scenarios/projections were already known in advance. These were:

1. the OEDC/IEA World Energy Outlook (IEA 2016d) 2. the US DOE/EIA Annual Energy Outlook (US EIA 2016)

Such scenarios present a weighted reflection of past experience, as well as the assumptions that they contain for the future. As such, they ‘digest’ historic price swings that can be extreme (and appear random or stochastic), to produce much smoother curves of future supply, demand and prices. The challenge is to correctly see with hindsight through the historical noise that obscures the underlying trends, and then to extend these trends into the future. It is always an exercise in humility to compare the jagged historical cost data with the smooth future curves, and to know “the forecast is always wrong,” but that significant lessons can also be learned about different future strategies, and their tradeoffs, risks and resiliency.

Oil, gas and coal prices have all declined significantly in the last several years, reflecting both the drop in demand due to economic slowdown (including China), and increased production capacity and expected reserves due to the revolution in fracking for tight oil and natural gas.

Darwinian competition in fracking for oil and gas has driven the technology forward, while increased Saudi production and dropping oil prices have left only the best surviving. This has provided the US oil and gas sectors with a base of known geology and already drilled wells (with well costs already liquidated by the failures of the weakest producers) ready to increase production with higher prices. As a result, discussion of peak oil fears has significantly subsided, and long-term price expectations have declined. This can be seen, for example, in (IEA 2016d), where an additional year of low energy prices for 2015 has slightly reduced the long-term projection slightly below the 2015 scenario. Nuclear fuel prices have also declined precipitously, reflecting a drop in fuel demand following Fukushima, as well as other factors (see below).

The main concern and effort of the fuel price survey was to find other long-term projections that would improve or supplement (IEA 2016d, US EIA 2016). The three main sources considered were: 1) other government or international agencies, 2) energy industry companies and associations, and 3) energy sector consultants and/or data service providers.

Individual sources within these groups are discussed below.

5.2.3.1.2 Other Government / International Agencies

Beyond (IEA 2016d, US EIA 2016), an effort was made to survey other government and international agencies. For the fossil fuel prices in particular, these included the following:

• United Nations (UN) – The UN has policy concerns and goals regarding energy, including economic development, sustainability and climate change. These are reflected through the UN-Energy interagency mechanism, the UN Environment Program, the UN

191

Foundation and other agencies. The UN Energy Statistics Division collects historic national data from more than 190 countries and maintains the Energy Statistics Database. However no actual energy price scenarios were found.

• European Commission (EC) – The EC performs energy modeling for the EU, and the EU Reference Scenario 2016 contains a graph of future fuel prices (but few numbers, or any tables of projected prices) (EC 2016a). However, the report states that the Reference Scenario 2016 is based on the PROMETHEUS model of international price projections, which is consistent with the New Policies Scenario of (IEA 2015g) for the medium and long term, and the WEO projections for all scenarios were already in hand.

• UK Department of Energy and Climate Change (DECC) – Fuel prices were taken from the DECC 2015 Fossil Fuel Price Assumptions (DECC 2015), which however states the caveat that the projections are “not forecasts of future energy prices.” This is typical of the reservations by modelers that scenarios are not the future, and there is no necessary probability associated with different scenarios.

• World Bank – Publishes the World Bank Commodities Price Forecast, including oil, natural gas and coal to the year 2025 (World Bank 2016).

• International Monetary Fund (IMF) – Publishes the IMF Commodities Price Outlook, but only to 2018 (IMF 2016).

It is possible that this survey may have missed some other government/agency fuel price projections (particularly possible non-English, national modeling efforts, such as China).

However, (IEA 2016d, US EIA 2016) both contain their own internal comparisons with other projections, and also do not mention any other major government or agency fuel price scenarios.

For nuclear fuel, the following government sources for price projections were reviewed:

• IAEA/OECD – The preeminent resource data source is the uranium Red Book (OECD/NEA/IAEA 2014, OECD/NEA/IAEA 2016) issued jointly by the International Atomic Energy Agency (IAEA) and the OECD Nuclear Energy Agency (NEA). The Red Book contains uranium resources by resource certainty, country, geology, and the method and cost to produce. It also contains two nuclear demand scenarios to 2035, but does not combine supply and demand data to quantify future prices.

• EC Euratom Supply Agency (ESA) – The ESA focuses on nuclear energy in the EU, including markets, EU supply and demand, security of supply, and medical isotopes (EC 2016b). It follows EU prices for 1) spot uranium prices, 2) the long-term price paid by utilities under multi-year contracts. Multi-year contracts are the dominant form of purchase (~90%), but the ESA only reports historical price data and does not report future prices on its website or in the ESA Annual Report. In general, “long-term” means

“multi-year contracts” in relation to nuclear fuel prices, and does not refer to price projections into the long-term future (e.g., 2040 or 2050).

• OECD/IEA WEO – The World Energy Outlook issued by the OECD’s International Energy Agency does not give prospective uranium prices as it does for fossil fuels. The OECD’s other primary scenario document is the Energy Technology Perspectives 2016 (ETP), which only refers to the Red Book.

• USDOE/EIA – The Annual Energy Outlook includes nuclear generation, but does not include future price data as it does for fossil fuels.

192

• DECC – The DECC Energy Markets Outlook (DECC 2016) focuses on the supply and demand of nuclear fuel in the UK and globally, but does not provide long-term prices, and is based on the Red Book and the NEA Nuclear Energy Outlook.

5.2.3.1.3 Energy industry companies and associations

The major oil and gas companies do their own future supply and demand modeling, and sometimes publish quite detailed supply and demand projections by region and demand sector. The modeling required to produce these projections presumably must also produce the prices necessary to balance the supply and demand, but the industry apparently does not believe that it is to their advantage to publish these. The oil and gas companies or associations surveyed included:

• OPEC – Tuture prices are published in and taken from the World Oil Outlook 2015 (OPEC 2016), which contains no future natural gas or coal price scenarios for comparison.

• British Petroleum (BP) – Detailed supply and demand projections are published in the BP Energy Outlook, 2016 edition, but no prices (BP 2016).

• Exxon Mobil – Publishes The Outlook for Energy: A View to 2040. Also includes detailed region and sector demand projections but no future price data (ExxonMobil 2016).

• Shell – Publishes the New Lens Scenarios: A Shift in Perspective for a World in Transition (Shell 2016), which describes two qualitative scenarios. The Mountains scenario envisages maintaining the status quo and stability with limited market forces and social mobility, while the Oceans scenario contains more devolved and competing interests with more economic productivity but possibly destabilized social cohesion and political stability. Future fuel prices are not quantified, but qualitatively the prices are

“moderate” and “higher” for the two scenarios, respectively.

• Eurogas – Publishes the Natural Gas Demand and Supply – Outlook to 2030, which gives supply and demand projections, but no prices (eurogas 2016).

Coal market price projections were sought, but the search results were generally either too short-term (circa 2020), or referred to other commodities price scenarios already mentioned ((IEA 2016d, IMF 2016, US EIA 2016).

In general, the industry survey of future fossil fuel prices shows that the strong emphasis is first on global crude oil prices, only second on natural gas prices and future coal prices are a distant third. Due to the goal of having coordinated, coherent price projections for all three fossil fuels, the industry sources are of less interest than the government/agency sources.

• World Nuclear Association (WNA) – The WNA website provides basic descriptive information about the nuclear fuel cycle, including historic uranium prices, historic cost supply curves, etc., but does not show fuel price projections. The WNA sells the Nuclear Fuel Report (WNA 2016a), including supply and demand scenarios to the year 2035, but does not provide any open future fuel prices.

• Nuclear fuel industry companies – Similar to the fossil industry, the nuclear industry fuel companies focus on either historical or short term future fuel price information.

5.2.3.1.4 Consultants and industry service providers

There is a very active industry of consultants and industry data providers that serve the fossil and nuclear fuel sectors. The problems here however are twofold. First, the commercial emphasis of the market means that there is a short time horizon, and second,

193

the data providers are very definitely for-profit, and serving a market that pays well. This means that any long-term, publically available price projections are rather meager. The following consultant/industry sources were surveyed.

• Deloitte – Deloittes’ Oil and Gas Price Forecast, September 30, 2016 (Deloitte 2016) contains oil and natural gas prices to 2035 for different benchmark types and/or locations.

• Bloomberg – The Bloomberg New Energy Outlook 2016 (Bloomberg 2016) focuses on new renewable energies and the power generation sector rather than fossil fuels.

• Lazard – Lazard’s Levelized Cost of Energy Analysis (Lazard 2016) focuses on forecasting levelized costs of competing conventional and renewable electricity generation technologies, with some fuel price assumptions, but does not contain any real long-term, comprehensive fuel price data.

• Wood Mackenzie – Wood Mackenzie is a well-established energy sector consulting company that produces its own commodity price projection, but the only publically available projection (Wood Mackenzie 2016) does not include prices.

A small number of price projections from the fossil sector were obtained by the fuel price scenario comparisons published within (IEA 2016d, US EIA 2016). Often these comparisons have been based on direct personal contacts (either by phone or email) or from sources that are not publically available, or are available only by purchase (generally at expensive consulting company rates). These sources include ArrowHead Economics (arrowhead 2016),

A small number of price projections from the fossil sector were obtained by the fuel price scenario comparisons published within (IEA 2016d, US EIA 2016). Often these comparisons have been based on direct personal contacts (either by phone or email) or from sources that are not publically available, or are available only by purchase (generally at expensive consulting company rates). These sources include ArrowHead Economics (arrowhead 2016),

Im Dokument Potentials, costs and environmental assessment of electricity generation technologies (Seite 186-200)