Economic Feasibility and Investment Decisions of Coal and Biomass to Liquids 1

Transcription

1 Economic Feasibility and Investment Decisions of Coal and Biomass to Liquids 1 Oleg Kucher, Ph.D. Student Natural Resource Economics Program, West Virginia University Morgantown, WV USA Phone , okucher@mail.wvu.edu Jerald J. Fletcher, Professor and Director West Virginia University US-China Energy Center, Morgantown, WV USA Phone , jjfletcher@mail.wvu.edu Abstract Coal and biomass to liquids (CBTL) technologies can produce synthetic fuel with lower CO 2 emissions than petroleum-based fuel (NETL, 2009). However, the economic feasibility of CBTL production depends on technological, economic and policy factors that may not be favorable in the U.S.; high capital costs of CBTL plant development is of particular concern. CBTL is an emerging technology; no CBTL plants have been constructed. This paper focuses on economic factors that determine the long-term economic feasibility of developing CBTL fuel in the U.S. The CBTL technology scenario draws upon a CBTL plant of 50,000 barrels per day (bpd) with a biomass input of 7.7% by weight designed by NETL (2009). We apply a micro-economic analysis to the projected CBTL plant operation by providing a discounted cash flow (DCF) analysis and real options valuation to assess the economic feasibility and investment decisions for the CBTL plant. Specifically, we incorporate uncertainties in fuel prices and capital costs in the DCF analysis and perform a Monte Carlo simulation to derive the volatility of the net cash flow from the CBTL plant. We then apply the derived volatility and estimates from the DCF analysis into a real options continuous-time model of irreversible investment to evaluate the CBTL investment options. The model results show that the value option to invest exceeds the net present value (NPV) from DCF analysis over four times and would tend to delay construction of CBTL plants in the near term. Our analysis suggests that lower capital cost and better economics of the CBTL project would improve the potential for investments into CBTL plants in the U.S. Keywords: Coal and biomass to liquids economics, investment, uncertainty, real options. 1 Acknowledgment: This material is based upon work supported by the Department of Energy under Award Number DE-FC26-06NT The authors would like to thank Thomas J. Tarka for valuable suggestions on the CBTL capital costs. Disclaimer: This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof. 1

2 1. Introduction Coal and biomass to liquids (CBTL) technologies can produce synthetic fuels such as diesel and jet fuel with lower CO 2 emissions than petroleum-based fuels (NETL, 2009). Recent studies suggest perspectives of CBTL production of affordable and low-carbon diesel fuel on a large scale in the U.S. (NETL, 2007, 2009; Bartis et al., 2008; Hileman et al., 2009; Darmstadter, 2010). In particular, a study by NETL (2009, p.51) underlines that CBTL approach enables the economic production of 20 times more diesel fuel from secure, domestic energy resources. The U.S. Department of Energy reports further financing of research in carbon capture and storage (CCS) technologies to reduce life-cycle CO 2 emissions by investing $4 billion in CCS with expected $7 billion from the private sector (U.S. EIA, 2010). At the same time, the economic feasibility of CBTL production depends not only on availability of technology but also on economic and policy factors that may not be favorable in the U.S.; high capital costs of CBTL plant development is of particular concern. CBTL is an emerging technology; no CBTL plants have been constructed. Beginning in 2011, only three CBTL projects have been considered for commercial deployment in the U.S. with a combined production capacity over 110 thousands barrels per day (bpd) of synthetic fuels. But all projects are delayed and only one of them could be possibly ready after It is not clear whether these projects are financially viable in the long-run due to uncertainty. Indeed, the efficiency of investment in CBTL assets in the U.S. cannot be predicted with confidence because of the indefinite capital costs, high volatility of energy prices and uncertain carbon requirements. In this paper, our main concern is high investment outlays for the new CBTL plant. Investments in CBTL plant possess two important characteristics: irreversibility and flexibility, described by Pindyck (1991) and Dixit (1992). First, the investment in CBTL plant could be a sunk cost if the project expenditure cannot be fully recovered due to uncertainty in future energy prices and costs. Second, the investment outlays have the opportunity value, since the firm has flexibility to delay investment until the better project conditions. These economic factors need to be included in the project economics in order to correctly assess the economic feasibility of CBTL. This paper focuses on the economic factors that determine long-term economic feasibility of developing CBTL in the U.S. We apply a micro-economic analysis to the projected CBTL plant operation by providing a discounted cash flow (DCF) estimation and real options valuation to assess the economic feasibility and investment decisions for a CBTL plant. After performing the DCF analysis, we identify the major sources of uncertainty that affect investment decisions to proceed with CBTL plant construction. We then incorporate uncertainties in the DCF analysis to use a Monte Carlo simulation for determining the payoff of the CBTL project and its volatility. Applying the derived estimates into the real options model, we assess option value to invest arisen from irreversibility of capital costs and the uncertainty over the future payoff from CBTL investment in the U.S. Our micro-economic analysis is based on the NETL (2009) techno-economic design of commercial CBTL 50,000 bpd projected plant with 7.7% by weight (wt. %) biomass. This CBTL plant configuration is considered to be one out of five CBTL options analysed by NETL (2009) as the most pragmatic solution to economic efficiency and carbon compliance. The motive to consider it because this CBTL configuration will meet the required greenhouse gas (GHG) emissions profile reduction up to 20% below than petroleum-derived diesel GHG profile at the minimum cost. We employ the technical production estimates for the base case of coal and biomass acquisition modelled by RAND Corporation, and the energy conversion criteria designed by the GREET model and EPA MOVES model (NETL, 2009). Lastly, we apply economic and financial techniques to provide micro-economic analysis for this CBTL plant in the U.S based on up-to-date economic and financial assumptions. The present study extends the previous NETL (2009) analyses on economic feasibility of the CBTL plant operation in the U.S. by introduction of real options valuation, while dealing with the uncertainty in fuel prices and costs. The specific objectives are to: provide economic assessment of coal-biomass to liquids by valuing cash flows and net present value of the prospective CBTL plant primarily in the U.S.; estimate the magnitude of option value to invest into the CBTL plant under uncertainty, and draw insights on the investment opportunities for the CBTL plant in the U.S. in the near term. 2

3 Results Techniques Objectives 2. Research methods and model The research framework explores the micro-economic analysis of a CBTL plant in the U.S. The research approach incorporates DCF analysis, risk assessment and real options valuation. It intends to value the CBTL economic viability and investment decisions under uncertainty over time. The basis for economic feasibility for this CBTL plant is determined by a firm s maximization objective function maximize the value of the firm. Here, the value of the firm is the present value of its expected free cash flow from the projected CBTL plant, discounted at the cost of capital. Hence, investment to build the CBTL plant happens if the project receives positive net present value at acceptable rate of investment return. However, because of the uncertainty, this investment rule may be wrong as the stochastic changes in the project value would result in overinvesting or underinvesting in the CBTL. Therefore it is important to correctly assess the CBTL investment value to avoid the economic loss of building the new large-scale plant. In this way, we employ the basic continuous-time model of irreversible investment, originally developed by Mcdonald and Siegel (1986), extended by Dixit and Pindyck(1994). Empirical application of the variations of this model for energy projects has been undertaken by Kiriyama and Suzuki (2004), Rothwell (2006), Yang et al.(2008), Blyth (2010). At the same time, the real options studies do not consider large scale unconventional liquids investment projects. We further apply the real options model to the large-scale CBTL project. The complexity of the CBTL valuation requires undertaking different methods and techniques. Figure 1, shown below, outlines the theoretical framework applied in this paper. It illustrates the main approaches, its objectives, techniques and expected results required to evaluate CBTL plant investment decisions. Figure 1: Research framework for the CBTL evaluation 1. DCF 2. Risk analysis 3. Real options approach Evaluation of CBTL Project by valuing cash flows and NPV Estimation of uncertainties in sensitive variables for the NPV Valuation of investment opportunity Linear deterministic valuation of the series of cash flows discounted at the end of the year Sensitivity analysis NPV, IRR Present value of net cash flows Sensitivity results The probability occurrence and approximation of the distribution parameters Monte Carlo simulation The distribution parameters The volatility estimates Dynamic programming Value of option to invest The first traditional method for valuing the CBTL project payoff in this paper is a DCF analysis. The static DCF determines whether the CBTL NPV is positive and whether investment return is satisfied after assessing all costs and revenues. As these measures are uncertain, the second method evaluates risks assessing the extent of uncertainties in economic variables. Risk analysis results are further implemented in a Monte Carlo simulation to simulate the expected present value of cash flow and its volatility. Finally, the real options analysis quantifies the flexibility of investment decisions from the immediate investment or wait resolving uncertainty Discounted cash flow analysis The DCF linear deterministic model is constructed to estimate the CBTL payoff by discounting the net free cash flow to present value. We estimate the free cash flow for a prospective CBTL facility, using concepts of assets valuation, free cash flow to firm (FCFF) and free cash flow to equity (FCFE) based on generally accepted accounting principles. The free cash flow comprises operational cash flow, capital expenditures, working capital and financing capital. The NPV is estimated as a sum of the present values of the expected net cash flow, and calculated as a function of operating revenues, operating expenses and costs, including operation and maintenance costs, taxes, interest rates and investment expenditures. The input-output physical parameters in the DCF model are taken from the NETL designed CBTL configuration plant (NETL, 2009) Risk analysis and Monte Carlo simulation In this paper, risk analysis associates with an uncertainty of the value of this project. The uncertainty of the payoff is derived from fitting the probability distribution of the sensitive variables and specifying correlations among those 3

4 variables (Belli et al., 1998). Specifically, we use the probability occurrence technique to assign the best-fitting probability distribution parameters based on historical data (Mun, 2006). We also refer to the recent research literature and professional reports to justify the probability assumptions and correlations among selected sensitive variables. Given the probability assumptions, we then assess the volatility of the net cash flow by running a Monte Carlo simulation of the underlying probability distribution of variables outcomes in the DCF analysis. The volatility of the expected present value of net cash flow is then used in the real options approach to determine the value of investment options Real options model In this section, we present a real options basic continuous-time model of irreversible investment after Dixit and Pindyck (1994). Here, we consider investment decisions for this CBTL project under uncertainty emphasizing the investment options opportunities through value option to invest and value option to wait. These values are based on the expected present value of free cash flow from the CBTL plant derived in the DCF analysis. The latter option suggests that firm can improve the investment conditions by optimizing project payoffs. Following Mcdonald and Siegel (1986) we assume that value ( ) follows a geometric Brownian motion (GBM) due to the uncertainty of the future cash flow, i.e. high volatility of energy prices, uncertain capital costs and uncertain regulation. In case of long-run energy prices, Pindyck (1999) suggests that GBM assumption is appropriate for many energy applications. The notation follows after Dixit and Pindyck (1994, chapter 5, sections 1, 2). Define the present value of the net free cash flow per year t from CBTL plant; present value of investments or capital expenditures in DCF analysis; discount rate; than value option to invest, can be defined as payoff from investing in the CBTL plant: (1) where is expectation and is the future time when investment is made into CBTL project. The equation (1) says that investor maximizes the payoff subject to value change of cash flows which follows GBM: where a drift parameter or growth parameter that increases at a factor of time, current value of the project with volatility; the increment of Wiener process, representing normal random variable to capture uncertainty. Since is stochastic we want to value an expected payoff from investing, and the critical value,, at which it is optimal to invest in the CBTL plant. In order to solve this investment problem, we follow dynamic programing with adding uncertainty, represented by volatility of the present value of the cash flow, with drift that creates a value of waiting. We assume a risk neutral investor, so that we can replicate component in dynamic optimization. The dynamic point is, that is the value of an option to invest into the CBTL project equals to an expected value of net cash flow. Because the investment opportunity, yields no cash flow until the time, the only return is its capital appreciation, which is represented through Bellman equation after Dixit and Pindyck (1994), as: The above equation says that over time, the return on investment equals to the expected rate of capital appreciation. We expand using Ito s Lemma, that shows first and second order differentiation of the value, as: Substituting GBM equation for and noting that, because is normally distributed with mean 0 gives: (2) (3) (4) (5) Substituting equation (5) into (3) and dividing both sides by and rearranging, we have: 4

5 Because of the stochastic changes in payoff we have to deal with two parameters: growth rate and volatility, For simplicity Dixit and Pindyck (1994) suggest a substitution for, where is dividend yield, assuming that, or. The dividend rate is viewed as the rate of a capital gain or an opportunity cost of delaying construction of the CBTL plant. Given this condition, the Bellman s equation becomes: (6) (7) In order to have solution, the value of investment opportunity must satisfy the boundary conditions: 1 (8) (9) (10) Equation (8) says that there is no option value of investment opportunity if net cash flow goes to zero. Equation (9) says that critical value of investment opportunity equals to net of the critical value of the present value of net cash flow and investment. In other words, value of the project equals to the full cost (opportunity cost + direct cost) of making investment: Obviously, the first order differentiation of the value equals to 1. Finally solving equation (7) subject to the boundary condition (8) yields the solution as following: where and positive constants to be determined which account for changes through volatility and cost of capital. The other important outcomes of these applications are trigger value,, and positive constants:, and, derived by substituting equation (9) into equation (8) and (9) and rearranging, as following: Equation (12) derives the trigger value of the investment option which makes investor indifferent between the investment right away and waiting. It enables to determine the value of waiting and the optimal solution. (11) (12) (13) To solve for of the solution the general solution to the second order differential equation (7) is expressed as a linear combination and by substitution it satisfies the quadratic equation as following: with the quadratic roots: with the general solution: (14) (15) (16) Since the boundary condition implies that, we end up with the equation (11). The solution is the value of opportunity to invest, that can be expressed as following: (17) 5

6 where From the mathematical expression in (17), we can see that value of option to invest, yields two different upward trends of the value of the project. The first equation represents the value option to wait, denoted as the value of the project, with the positive slope and with an exponent. The second equation for the NPV of the CBTL plant gives the traditional NPV. With the inequalities on the right hand sides, this equation modifies the optimal investment rule: invest if the value of the project is greater than the trigger value, otherwise wait. So, if the value of the project is less or equal than the trigger value, the investment should be postponed. If the value of the project is greater than trigger value,, than investor is rather to invest into the project. Positive constant depends on three variables: discount rate,, dividend yield,, and volatility,. The latter is the standard deviation of the present value of the net free cash flow. Positive slope depends on volatility of the project value and investment outlays. The threshold,, serves as the critical point at which investor decides whether to invest or wait until more information would be available. Mathematically it depends on investments, and an expression with. The expression depends on uncertainty in the way that greater volatility results in a higher value option to invest. Contrary, higher discount leads to smaller constant, meaning that the more we discount our future value, the less would be option value. In order to calculate the value of investment option under uncertainty we start to evaluate the option of building the CBTL 50,000 bpd plant with 7.7wt% biomass in the U.S. in the next few years with starting the commercial operation in A section 3.1 provides the results of DCF and sensitivity analysis for the CBTL plant with estimating the present value of cash flow and NPV. Section 3.2 estimates the mean and the volatility of the present value of the cash flow using the Monte Carlo simulation based on the distribution assumptions of the key DCF variables. Section 3.3 presents real option valuation results built on the DCF results and the estimated uncertainty of the CBTL project value. 3. The CBTL project evaluation To evaluate the CBTL plant we construct a comprehensive dataset of costs, revenues, and other economic parameters using the best available information from the NETL, RAND latest reports and investors data sources. Several complex assumptions have been made about the different project parameters with respect to size, plant life, construction period, capacity factor, unit prices, escalation rate, cost of capital product, tax rates, depreciation, working capital, interest during construction, and other parameters of a plant (appendix A). Note, the CBTL project evaluation is based on the real prices and unit costs as of 2010 in the energy market. Free cash flow to firm is considered as the base case for real options valuation with the procedure utilized in appendix B. Applying this dataset to the real options model, we draw the conclusions on the value of option to invest. Firm-level analyses add caveats but show the realistic implications for CBTL liquids development. The DCF analysis of this CBTL project is performed using self-constructed MS Excel spreadsheets. The analysis of capital costs, depreciation and financing costs over the plant lifetime is conducted on NETL Power Systems Financial Model 6.6 (Worhach, 2011). We fit the probabilities for the sensitive fuel variables and derive their parameters based on the historical data, followed by performing a Monte Carlo Simulation for the free cash flow using the Risk Simulator (Mun, 2006) DCF analysis The discounted cash flow analysis estimates potential future payoff from the CBTL project during the life of the project. The life of the project consists of the construction period for the CBTL plant at 4 years and operating life for the CBTL production at 30 years with average operating capacity factor at 89% (appendix A). 6

7 Valuing cash flow of the CBTL project Cash flow is primarily determined in the form of operating revenues from selling CBTL output products in the market and costs from the purchase of input necessary for production and other operating expenses (i.e. operation and management (O&M) cost, taxes, interest etc.) for each year over the life of the project. For the input, we estimate coal, biomass, water, chemicals and slag expenses; for the output sales of diesel, naphtha and sulfur based on their prices obtained from the U.S. DOE (appendix A). The operating costs, output unit prices are escalated at 2% rate annually (NETL, 2009). The estimation procedure of the free cash flow is presented in appendix B. Overall, the CBTL plant is cash intensive project. The average operating revenue from CBTL is calculated at about $1.87 billion dollars per year, including 78% of revenue from diesel and 22% of revenue from naphtha (figures C2, appendix C). The average total operating expenses is estimated at about $813 million dollars per year if operating at 89% capacity. The patterns of cash flow, operating revenues, and costs for the CBTL plant are presented in figures C1 and C2 in appendix C. Note that operating expenses is the major part of total operating costs 60%, including coal costs 48% and biomass costs 8% (figure C1, appendix C). Capital costs reflect the value of the CBTL plant s investment outlays, I, expressed as total as spent capital (TASC). We follow the NETL procedure to estimate TASC cost. The total as spent capital is estimated at $5.6 billion (table 1). The TASC costs consist of engineering, procurement, and construction (EPC) cost (94%) and interest during construction (6%). The bare erected cost is also included in TASC costs comprising a plant equipment costs and expenses on facilities and infrastructure. The EPC cost includes the cost of EPC services and contingencies. Table 1: Capital costs of the CBTL 50k bpd plant with 7.7wt% biomass Capital costs (in million dollars) Value Percentage Plant equipment costs Coal & biomass handling preparation and feed $760 14% Gasification $1,476 26% Air separation, syngas cleaning & shift $728 13% Fischer Tropsch (FT) synthesis $725 13% CO 2 removal & compression $270 5% Other( boiler, steam turbine, ash handling, accessory electrical plant, instrumentation & control, site improvements & buildings) $592 11% Bare erected cost (BEC) $4,650 83% Engineering, procurement, and construction (EPC) cost $4,974 89% Contingency $275 5% Total capital costs $5,249 94% Interest during construction (IDC) $346 6% Total as-spent capital (TASC) $5, % Note: Based on the NETL (2009) projected plant scenario: CBTL bpd, 7.7wt% biomass. The plant equipment costs for the coal and biomass handling, preparation and feed, gasification of the feedstock and FT synthesis accounts for more than half of the total capital costs. The breakdown of EPC cost includes: gasification cost 30%, FT synthesis 16%, coal and biomass handling, preparation and feed 16%, air separation, syngas cleaning and shift 15%, boiler, steam turbine, ash handling and accessory electrical plant cost 6%. The CO 2 CCS facility cost is about 6% and the rest cost is estimated at 11% of plant cost. The EPC cost is escalated at 20% to yield the total capital costs. Note, the U.S. DOE reports 10% increase in capital costs for electric power plants (U.S. EIA, 2011). In sum, building the projected CBTL plant could potentially cost over $5.6 billion. Given the estimates of operating revenues, operating costs and capital expenditures and assuming tax rate at 38%, we value the average free cash flow from the operations of the CBTL plant at about $191 million dollars per year to establish a base case for further assessment. Note, that depreciation and working capital are insignificant determinants as depreciation serves as non-cash items while change in working capital from year to year is assumed to be relatively low. 7

8 NPV NPV Deterministic NPV, IRR The traditional net present value from the free cash flow to the CBTL plant discounted at 8% (NPV8) is estimated at $767 million dollars. We also report results for the NPV from the equity investors standpoint. Here, the NPV from free cash flow to the equity discounted at 12% (NPV12) is estimated at $207 million dollars over Such a difference is explained by different methods of estimation, different discount rates and assumed structure of financing capital. Internal rate of return (IRR), known as expected economic rate of return, from the pre-debt, after-tax cash flow of the CBTL plant is 9.2%, which is lower than IRR for equity investors of 12.9%. NPV and IRR profiles for the CBTL plant are illustrated in figure 2. Return on investment for the free cash flow to the CBTL plant is 15.4%. Figure 2: The NPV at discount rate for the CBTL plant over 30 years Free Cash Flow to Firm Millions of dollars $14,000 $12,000 $10,000 $8,000 $6,000 Internal Rate of Return $4,000 $2,000 $0 -$2, % -$4, % 4.2% 8.4% Cost of Capital, % 12.6% Millions of dollars $12,000 $10,000 $8,000 $6,000 $4,000 $2,000 $0 -$2,000 Free Cash Flow to Equity Internal Rate of Return on Equity 12.9% -$4, % 4.2% 8.4% 12.6% Cost of Equity, % The probability index is 1.6 from FCFF and 1.4 from FCFE, meaning that in money terms the investors will earn these amounts on every dollar invested in the CBTL plant. The payback period is 13 years to the firm and 15 years for the equity investors of recovering investments based on the cash flow and assumed cost of capital. According to the traditional DCF analysis, these results are consistent with the investment decision to invest into the project. However, this might not be true if uncertainty is taken into consideration. These static estimates provide several insights into the CBTL economic feasibility and investment decisions under traditional DCF. First, the positive NPV of the CBTL plant suggests the acceptance of this project, taking discount rate, inflation and other economic assumptions into consideration. Second, the IRR, as cost of capital to the firm and the cost of equity to equity investors, are relatively low, correspondingly at 9.2% and 12,9%, suggesting that such cost of capital won t be sufficient for the large-scale newly CBTL plant construction. Third, the patterns of NPV at cost of capital and equity indicate relatively high sensitivity of NPV, such as the higher the slope of NPV the greater its sensitivity to the change of the discount rate Sensitivity Analysis Sensitivity analysis assesses the extent of the influence of the DCF variables on the NPV and IRR. By conducting sensitivity analysis, we identify the most critical variables and their effects on the economic payoff by changing variables (± 10% of the base values) and deriving their effects on the NPV, using Risk Simulator software (Mun, 2006). Figure 3 shows the most sensitive variables in descending order for the change in NPV8 for the CBTL plant. The highest effects on the change in NPV8 (the upper level of the chart) are attributed by: fuel prices, operating capacity rate, discount rate, and investment outlays. Note, right bars indicate a positive effect on the NPV; while left bars show a negative effect in corresponding to the increase (green bar) or decrease (red bar) in the parameter. Therefore, for investment the red bar on the right side indicates that the lower the capital cost (total as-spent capital cost) the higher NPV; in other words, the investment and NPV are negatively correlated. Similarly, discount rate, tax rate, and input prices negatively affect the NPV. The opposite is true for the prices of fuel products. 8

9 Figure 3: Impacts of major sensitive parameters ( 10 %) on the NPV8 from the FCFF NPV Millions of dollars The broader picture of the sensitive variables and their quantifying effects to the NPV8 from the FCFF base value of $767 million is presented in the table C1, appendix C. The impact of the main 20 variables (± 10% change) is measured the upside or downside actual estimates of the change in NPV8 for the CBTL process. An increase by 10% in capital costs for the CBTL plant will yield three times lower NPV8 at $270 million. The spider chart illustrates the static impacts of major parameters on NPV of FCFF (figure C3, appendix C). Percentage change on each of the precedent parameters on NPV could be considered through the positively or negatively sloped lines represented by positive and negative effects consequently. Price of U.S. Ultra low sulfur diesel (ULSD) has the highest positive magnitude of a slope, and so its percentage change would have the highest effect on NPV. Contrary, the discount has the most negative slope, indicating that one unit increase of discount rate would decrease the NPV by the highest value, vice versa. Note, a 10 % change of discount rate is ranged from 7% to 9% with its value at 8%. Similar results for the NPV12 from the free cash flow to equity are shown in figure C4, appendix C. Several financial variables, such as debt ratio and structure, interest and loan repayment would considerably affect the NPV. Note, the investment outlays will play the major role for the NPV from the equity standpoint Risk assessment and Monte Carlo simulation The risk assessment in this section intends to capture the fluctuations in the sensitive variables around the average and their likelihood of the realized values while accounting for correlation. We use the Risk Simulator software to fit the distributions for available historical data of the selected variables (Mun, 2006). We consider the most sensitive variables: fuel prices, operating capacity and capital costs to derive the volatility of the net cash flow. As part of the discount rate, we approximate a dividend rate. We do not consider less sensitive variables that cause low risk. Also, the input-output product estimates are taken as fixed. In fact, the operating capacity incorporates the risk exposure of the production change. For those variables where data is not available we use approximation. Data for the petroleum refinery utilization rate serves to derive an expected capacity factor of this CBTL plant scenario. For the dividend rate, we use data of 10-Year Treasury constant maturity rate. Variables, historical data description, distribution parameters and assumptions are given in table 2, figure C5 in appendix C, and characterized below. ULSD prices. CTL literature reflects strong connection of diesel prices to the prices of crude oil. Here, we consider ULSD prices separately from the crude prices because the variability in diesel prices is distinguished from the change in oil prices. The historical data for ULSD prices is only available from Due to this limitation we use 9

10 both the monthly data of U.S. No 2 ULSD less than 15 ppm wholesale/resale prices by all sellers and annual average on-highway real diesel prices in dollars per gallon, compiled from the U.S. DOE EIA. The fitting distribution for the available data is lognormal 3 with a likely sigma in the range of (figure C5, appendix C). Positively skewed data exhibit a greater likelihood of higher prices. For consistency we assign similar distribution assumptions. Table 2: Fitted distribution parameters and distribution assumptions for sensitive CBTL variables Variables Base values Distributional assumptions U.S. Ultra low sulfur diesel price, $/gallon $2.31 Lognormal 3: Mean St. Dev Coal price, $/t $44.6 Lognormal: Mean St. Dev Crude oil price$/ oil barrel (bbl.) $79.4 Lognormal: Mean St. Dev Operating capacity 0.89 Gumbel Minimum: Alpha Beta Dividend rate,% 6 Triangular: Mean 6 Min 10%, Max. +10% Capital costs, millions of dollars $5,595 Triangular: Mean 5,595 Min 10%, Max. +15% Crude oil prices. Here, the crude oil prices are tied only to naphtha prices (appendix A). It is observed, that the longrun crude oil prices are highly volatile (Pindyck, 1999). Historically, the imported real oil prices fluctuate continuously up and down. Using the real prices of imported oil from , compiled by the EIA, we can see different price trends: an increasing trend during with average annual increase of about 29.4%, then a declining trend from with an average yearly decrease of 7.5%, and the successive increasing trend again on average of 21.1% by The EIA forecast over suggests that the price of crude oil could be in the range of $50 to $200 dollars with a mean of $125 dollars per barrel in 2009 dollars (U.S. EIA, 2011). Performing distributional fitting of the real crude oil prices suggests the volatility of $34.2 dollars in real prices (figure C5, appendix C). Similarly, we assume lognormal distribution with volatility of $35 dollars from the expected mean. Coal prices. Historical data of bituminous coal real prices during , compiled by the EIA, suggest that coal prices fluctuate with about half as much the average oil price fluctuation. The EIA projects the coal prices will rise with an average increase of 0.2% per year up to 2035 period, and their forecasted range of low and high coal costs are negative -38% and positive 65% from the reference coal prices ($/million Btu) projected in 2035 (U.S. EIA, 2011). After fitting the lognormal distribution for the real prices of coal from 1949 to 2009 (figure C5, appendix C) we assume lognormal distribution for the coal prices with standard deviation of $12.45 dollars for the CBTL project. Refinery capacity rate is the second largest sensitive variable in the sensitivity analysis. But because there is no CBTL plant, we use a proxy of petroleum refinery utilization rate, which is calculated by dividing gross input to distillation units by the annual average capacity. Since refinery utilization rate utilizes supply constraints and demand change, the capacity rate is not stable over time. We use the EIA data for refinery utilization rate from 1949 to 2009 to fit the negatively skewed Gumbel minimum distribution with higher probability of higher number (figure C5, appendix C). The sixty one-year utilization rate is about 88%, with a standard deviation of 6.3%, which means that two-thirds of the time the utilization rate has been fluctuated between 82% and 94%. Given the fact that the CBTL is newer technology, we assign similar distribution with the mean of 89% and standard deviation of 6.3%. Capital costs. Public information regarding capital costs for CBTL facilities is not fully available. From the available information for the capital cost for a CBTL plant is roughly estimated to be: U.S. $5.5 billion for BPD CBTL plant at Wellsville, Ohio (Baard Energy, 2007); U.S. $4.5 billion for BPD BCTL facility with at least 5% (as measured by energy content) of biomass and with CCS of Natchez Project (Rentech, 2008). In research literature capital investment coal for the CBTL facilities could be summarized as follows: U.S. $1.1-$1.4 billion for BPD CBTL facility with 15% biomass (Hileman et al., 2009); U.S. $5.7 billion or $113,500 per daily barrel for a CBTL plant with 8% switchgrass (NETL, 2009); U.S. $6.5 billion for BPD CTL plant with CCS (NETL, 2011). 10

11 Given this variability, we assume triangular distribution similar to Mantripragada and Rubin (2011). Note, the maximum investments outlay is relevant to the recent NETL estimations of the capital cost for the CTL plant. The minimum outlays could be possibly due to the anticipated technological improvement. Dividend rate. As part of the cost of capital, the dividend rate plays an important role in real option analysis. It is a proportional rate in the total return rate on the assets, similar to the return on the stock. The commonly accepted long-term return on common stock in the last century is about 6% (NETL, 2009). A very high approximation for the dividend rate in this study is assumed to be 10-Year Treasury constant maturity rate, which is a common benchmark long-term risk-free rate. Historical data for the average annual 10-Year Treasury constant maturity rate suggests the moderate fluctuation in the rate in the long-term period. Therefore, we assume triangular distribution with a mean of 6% and ±10% change around the mean. Note, the dividend rate is the part of the discount rate of 8% in DCF analysis. After identifying the probability assumptions, we further proceed with specifying the covariances among the fuel variables in order to avoid errors in a Monte Carlo simulation. It is widely discussed that energy prices tend to follow the oil prices. Over the last decade diesel-prices were highly correlated with oil prices. A recent study by Oliver Wyman (2010) reports correlations of the price levels of crude oil and diesel at While this is true for crude and diesel prices, the coal prices exhibit the different trend than oil prices movement in some years due to the difference in demand and supply fundamentals for these commodities in the energy market. Pragmatic levels of the correlations between fuel prices are imposed in this study. We assume that coal prices will follow crude oil prices during the next 30 years with the correlation coefficient of 0.6. The diesel prices are expected to correlate with oil prices with correlation coefficient of 0.9 over the life of this CBTL project. The same assumptions are incorporated in correlation coefficients linking the pricing of coal and diesel products. Distribution assumptions from table 2 are further implemented in a Monte Carlo simulation to derive parameters for real options valuation. Specifically, we simulate the present value of the net free cash flow thousands of times by using the specified probability distributions of the uncertain combined project components in order to determine the volatility of net cash flow over time. It constructs the probabilistic scenario of forecasts of the expected value of CBTL project with the corresponding descriptive statistics, including probabilities, standard deviation, variance etc. Figure 4 shows that NPV8 quite likely falls between negative -$0.85 billion and $1.72 billion dollars given the 90% confidence interval. The mean of NPV8 is expected at about $497 million. Figure 4: The CBTL NPV8 (in millions of dollars) forecast after 4000 simulations Mean NPV8 $ million; standard deviation $780.4 million The right panel of figure 4 shows that there is over 25% probability that construction of the CBTL plant will yield negative NPV under assumptions in table 2 and appendix A. The expected mean of IRR is 8.7% (figure C6, appendix C). 11

12 This CBTL project would likely have negative expected NPV12 from free cash flow to equity at negative -$1 million dollars after 4000 simulations with mean of IRR on equity at about 12% (figure C7, appendix C). Simulations of fuel prices, capacity factor, capital costs and payout rate are simulated to determine the forecast of the variance of net free cash flow over time. Figure 5 presents the results of 4000 simulations. Average present value of net cash flow is about $185 million dollars per year with a standard deviation of about $24 million dollars. Figure 5: The CBTL present value of the net cash flow (NCF) forecast after 4000 simulations Mean present value of NCF, V $185 million; standard deviation $25 million Statistics Mean Median Standard Deviation 24.3 Variance Coefficient of Variation 0.13 Maximum Minimum 79.7 Skewness Millions of dollars Kurtosis 0.17 Percentage of Error Precision 0.42 at 95% Confidence A volatility of the net cash flow is represented by the standard deviation from the Monte Carlo simulation. With a standard deviation of $24.29 million dollars the coefficient of variation is about 0.13, meaning that the relative standard deviation of the net cash flow is 13%. Note, the volatility derived from lognormal expression (standard deviation of percentage change) of present value of net free cash flow from the Monte Carlo simulation is 9.4%. In summary, the simulation results illustrate likely understatement of the CBTL project deterministic NPV and net free cash flow if risk exposure is ignored. Given the assumptions in section 3.2, future payoff from the CBTL project falls on the mean value at about $497 million dollars, which is 35% as lower as the NPV8 of the CBTL plant in traditional DCF analysis. This lower NPV value is still positive to proceed with the investment Real options application for the CBTL plant In this section we use results from the DCF analysis and Monte Carlo simulation to estimate the real options value to invest in the CBTL plant in the U.S. The parameters for real options model are presented in the table 3. Table 3: Real options model parameters Parameters Base values, discount rate 0.08, dividend (payout) rate 0.06 present value of capital costs (TASC) millions of dollars $4,972.6 net present value of net cash flow to the firm millions of dollars $5,739.6, millions of dollars $ volatility of average present value of net cash flow, % The volatility of the present value of net free cash flow (NCF) from the CBTL is considered at 13.4% after 4,000 simulations represented in figure 5. This is the relative standard deviation of the value of the project, derived as:. Note, that the estimated volatility reflects only the fuel prices volatility, capital costs volatility, capacity factor change, and the uncertainty in the dividend rate, δ. The latter is linked to 10-Year Treasury constant maturity rate, and corresponds to a rate of 6%. The combined volatility parameter does not incorporate a carbon regulation uncertainty or technical uncertainty. With a variance of 1.8%, dividend rate of 0.06 and discount rate of 0.08,, and are estimated as following: 12

13 NPV 2.434, , $ 8, million. Considering all the various distribution assumptions on the volatility of the sensitive variables and parameters presented in the table 3 together, the value of option to invest demands higher level of the efficiency of the CBTL plant, lower level of capital costs and better payoff (figure 6). Figure 6: Value of the investment opportunity for the CBTL plant in the U.S. for =13.4%, =0.06, =0.08 Millions of dollars F(V) $5,000 Linear NPV. F V.. F V. $3,468 $3,000 $1,000 NPV -$1,000 I V* V -$3,000 -$5,000 $0 $1,000 $2,000 $3,000 $4,000 $5,000 $6,000 $7,000 $8,000 $9,000 $10,000 Present value of net free cash flow, millions of dollars Figure 6 indicates a value of option to invest for the CBTL plant in the U.S. The vertical axis is the expected payoff from investing in the CBTL plant, and the horizontal axis is the present value of the net cash flow from the CBTL plant. The blue line is divided by tangency point of trigger value and critical NPV into two parts: dotted line which shows the value option to wait, denoted as, and the continuous blue line that represents the value to invest, denoted as. The linear dark red line represents the projected NPV of the project, defined as where is the perpetual cash flow and is fixed investment. The marked dotes on this line from bottom to top are zero NPV under initial investments from the DCF analysis, deterministic NPV from the DCF analysis at about $767 million, and the value of investment opportunity. In summary, the value of the option to invest is about $3.5 billion. Up to this optimal value, the investment in the new CBTL plant should be postponed if the payoff of the project is below the threshold of about $8.4 billion, which is highly dependent on the cost of the investment,. Correspondingly, the CBTL plant should be constructed if the payoff exceeds the determined trigger value with consideration of volatility, dividend and discount rates. Even though the deterministic NPV from DCF analysis creates positive value to invest, the CBTL project should be delayed until the expected payoff would exceed the traditional NPV over four times. 4. Summary and policy implications CBTL technologies are viewed as a promising alternative to produce affordable transportation fuels from the domestic feedstock with a compliant level of CO 2 emission. However, despite the technical feasibility of CBTL processes, there is little evidence of strong commercial viability of the CBTL plant in the U.S. in the near-term under uncertainty in the fuel prices and costs. Our analysis shows that in the presence of uncertainty over the payoff from investing, high capital costs of the CBTL plant in the $5 to $6.5 billion range is the main barrier for the construction of the first large-scale CBTL plant in the U.S. The volatility in the future net cash flow of the CBTL project creates high value to wait contributing to delay investments in the construction of the CBTL plant. 13

14 Our DCF analysis suggests that construction of the new CBTL 50 thousands bpd plant is expected to yield a positive NPV of $767 million under stated assumptions with the investments in this plant of about $5.6 billion in nominal value. But does this payoff make CBTL feasible in the U.S. in the near term under uncertainties in energy market? We find that uncertainties in the fuel prices and capital costs could lower the payoff from the CBTL project by 1/3 rd. When we further incorporate the utilized volatility of the net free cash flow in the real options valuation model, the considered CBTL project will gain substantially higher value of option to invest - over four times as large as the NPV calculated using the DCF analysis. The option value of waiting could potentially reach 2/3 rd value of the initial capital costs for the new CBTL plant. Regardless of the positive NPV of the CBTL project, the real options analysis suggests to defer investments into a large-scale CBTL project in the U.S. until economic conditions are better. Another possible explanation of investments delay in a CBTL plant, which is not captured in our analysis, is carbon regulation uncertainty. The constraints associated with the uncertainty of carbon regulations will most likely drive the capital and operating costs up and worsen financing options for the new CBTL plant unless regulatory certainty of CO 2 sequestration would be achieved. In the long run, the CBTL technology development could lower the capital costs for the CBTL plant and offset the value of waiting. Thus, an important addition to our analysis would be the real options model extension to account for carbon regulation applicable across all fuel types produced in the energy market with an empirical estimation of option value to invest into a CBTL project. We must consider the results of our model in moderation due to the following limitations. Our analysis is based on technical and economic assumptions for the CBTL plant evaluation. The real options model in turn relies on the determined volatility, fixed dividend rate and the discount rate that highlight the payout of the project. It is challenging to precisely evaluate the economic feasibility of the CBTL project. While our analysis is affected by the set of assumptions, our results point to a considerably high level of the option value to invest into the new CBTL plant suggesting the necessity of improving the project economics and financing options in order to make the CBTL feasible in the U.S. in near term. Overall, our micro-economic analysis provides insights into the economic feasibility and investment decisions flexibility for the projected CBTL plant. The CBTL technology will need to be substantially more cost effective, either through reductions in capital costs or increased policy incentives, such as low-carbon subsidies or carbon taxes in order to make CBTL investments attractive and profitable. Unlike coal-to-liquids technologies, which have been commercially deployed in South Africa and China, commercialization of the CBTL plant in the United States faces economic hurdles and quite likely would be further deferred in construction in the near term. Advancing CBTL technologies to the level when it would be commercially viable could be possible through improving economics by reducing capital costs depending on optimized configuration of the CBTL plant, and increase in the CBTL product demand and the government support to attract the investment. 14

15 Reference Aswath, Damodaran Applied Corporate Finance: A User's Manual. John Wiley & Sons, Inc. Baard Energy "Ohio River Clean Fuels. Project Summary." (accessed August 26, 2011). Bartis, James T.; Frank Camm and David S. Ortiz "Producing Liquid Fuels from Coal: Prospects and Policy Issues." RAND Corporation, MG-754-AF/NETL. Santa Monica, California. Belli, Pedro; Jock Anderson; Howard Barnum; John Dixon and Jee-Peng Tan "Handbook on Economic Analysis of Investment Operations." Operational Core Services Network, Learning and Leadership Center. Blyth, William "The Economics of Transition in the Power Sector " OECD/IEA. Darmstadter, Joel "The Prospective Role of Unconventional Liquid Fuels." Resources for the Future, Dixit, Avinash K "Investment and Hysteresis." Journal of Economic Perspectives, 6(1), Dixit, Avinash K. and Robert S. Pindyck Investment under Uncertainty. NJ.: Princeton University Press. Duffy, Mike "Estimated Costs for Production, Storage and Transportation of Switchgrass." ISU. Hileman, James I.; David S. Ortiz; James T. Bartis; Hsin M. Wong; Pearl E. Donohoo; Malcolm A. Weiss and Ian A. Waitz "Near-Term Feasibility of Alternative Jet Fuels." RAND Corporation and Massachusetts Institute of Technology. Kiriyama, Eriko and Atsuyuki Suzuki "Use of Real Options in Nuclear Power Plant Valuation in the Presence of Uncertainty with CO 2 Emission Credit." Journal of Nuclear Science and Technology, 41(7), Mantripragada, Hari Chandan and Edward Rubin "Techno-Economic Evaluation of Coal-to-Liquids (CTL) Plants with Carbon Capture and Sequestration." Energy Policy, 39, Mcdonald, Robert and Daniel Siegel "The Value of Waiting to Invest." Quarterly Journal of Economics, 101(4), Mun, Johnathan Modeling Risk : Applying Monte Carlo Simulation, Real Options Analysis, Forecasting, and Optimization Techniques. John Wiley & Sons, Inc., Hoboken. National Energy Technology Laboratory "Affordable Low-Carbon Diesel Fuel from Domestic Coal and Biomass." DOE/NETL-2009/ "Cost and Performance Baseline for Fossil Energy Plants Volume 4: Coal-to-Fischer-Tropsch Liquids Using a Dry-Feed Gasifier." (Draft 8/8/2011) "Cost and Performance Baseline for Fossil Energy Plants, Volume 1: Bituminous Coal and Natural Gas to Electricity." DOE/NETL-2010/ "Increasing Security and Reducing Carbon Emissions of the U.S. Transportation Sector: A Transformational Role for Coal with Biomass". DOE/NETL-2007/ "Recommended Project Finance Structures for the Economic Analysis of Fossil-Based Energy Projects." DOE/NETL-401/ Oliver Wyman "Food/Fuel Price Dynamics: Developing a Framework for Strategic Investments." (accessed August 26, 2011). Pindyck, Robert S "Investments of Uncertain Cost: An Application to the Construction of Nuclear Power Plants," E. S. Schwartz and L. Trigeorgis, Real options and investment under uncertainty: classical reading and recent contributions. Cambridge: The MIT Press, Pindyck, Robert S "Irreversibility, Uncertainty, and Investment." Journal of Economic Literature, XXIX, Pindyck, Robert S "The Long-Run Evolution of Energy Prices." The Energy Journal, 20(2), Rentech, Inc "Synthetic Fuel Initiatives." 3rdAnnual Coal-to-Liquids & Gas-to-Liquids Conference. (accessed August 26, 2011). Rhodes, James S. and David W. Keith "Biomass Co-Utilization with Unconventional Fossil Fuels to Advance Energy Security and Climate Policy." Washington DC. Rothwell, Geoffrey "A Real Options Approach to Evaluating New Nuclear Power Plants." 27 (1), United States Energy Information Administration "Annual Energy Outlook 2011." DOE/EIA-0383(2011). United States Energy Information Administration "Meeting the Energy and Climate Challenge", 2010 Energy Conference: Short-Term Stresses, Long-Term Change. Washington, DC. Worhach, Paul (Nexant) "Power Systems Financial Model Version 6.6 User's Guide". DOE/NETL- 2011/1492. Yang, Ming; William Blyth; Richard Bradley; Derek Bunn; Charlie Clarke and Tom Wilson "Evaluating the Power Investment Options with Uncertainty in Climate Policy." Energy Economics, 30(4),

16 Appendix A: Parameters and assumptions utilized in the DCF analysis of the CBTL plant Items/Description Parameters Assumptions Plant basis: Type plant: CBTL 50k bpd with The CBTL plant design is drawn from (NETL, 2009). 7.7% biomass by weight Life of project: 30 years The life of the project is assumed to be 30 years (NETL, 2009). Construction period 4 years The plant construction is assumed to be completed prior in 4 years The base year 2012 The base year for the CBTL plant construction is The start-up year 2016 The startup-year for the CBTL plant production is Days of operation: 325 days per year The operating period is assumed as following: year one 10% of maximum operation days, year two 70%, year three is 80%, year four 90% of maximum operation days per year. The maximum production will occur in the fifth year. Operating capacity 0.89 The designed capacity factor of 89% is achieved in the fifth year of operation. * Major Input: Quantity * Unit Price Coal feed, tpd $44.6 The coal price is the delivered price of coal for electric power in the U.S. in 2010 (2009 US$ per short ton) from the U.S. DOE EIA AEO 2011 Table Coal Supply, Disposition, and Prices, Reference Case (U.S. EIA, 2011). Biomass feed, tpd 1657 $90.0 The biomass feedstock cost is assumed to be roughly twice the cost of coal (Rhodes and Keith, 2010). Note, Duffy (2007) estimates the total switchgrass cost from $80 to $113.6 per ton based on various yield per acre in Water, thousands gpd $1.08 The water cost is the cost per 1000 gallons (NETL, 2011). Chemicals, t/d 30 $1000 The chemical cost is the cost per 1 ton (NETL, 2011). Slag, t/d 1909 $16.2 Slag disposal costs is $16.23 from the IGCC case (NETL, 2010). Major Output: Quantity * Unit Price Diesel, bpd The diesel price is assumed to be $2.314 per gallon of the U.S. N 2 diesel retail sales by refiners in 2010 (U.S. EIA, 2010). We multiply diesel price by 42 to value ULSD price per barrel. Naphtha, bpd The naphtha is valued at 77% of the petroleum crude (NETL, 2007). We use the crude oil spot price of West Texas Intermediate at $79.4 dollars per barrel in Sulfur, tpd The average price for sulfur is assumed to be $36.29 per ton as of 2007 from the U.S. Geological Survey. Financial Parameters: Discount rate, % 8 The discount rate of 8% exceeds 7% rate from NETL (2009). The risk free-rate of 6% is assumed as part of the discount rate. Inflation, % 2 The average annual inflation rate is assumed at 2%. Note, GDP price index is forecasted at 1.8%, while consumer price index for energy commodities at 2,9% for (U.S. EIA, 2011). Income tax rate,% 38 The income tax rate is assumed to be 34% federal and 6% state. Interest rate, % 5.5 Interest costs during construction are calculated in the PSFM 6.1. Initial working capital 7 The initial working capital is set at 7% of first year of revenues. Debt equity ratio,% 60/40 The debt financing structure for the CBTL project is assumed to be 60% debt and 40% equity (NETL 2008, 2009). Debt structure, % 80/20 A financial debt structure is assumed at 80% of a senior debt for 20 years and 20% of a subordinated debt for 15 years. Senior debt interest,% 5.5 The senior debt is loan guaranteed with 5,5% interest. Subordinated debt 9 The subordinated debt interest rate is assumed to be 9%. interest rate, % Depreciation schedule, years 20 The depreciation of capital costs is calculated over 20 years period by using 150% declining balance method. Capital cost escalation, % 20 The EPC cost from the NETL (2009) study is escalated by 20%. Note: * The quantity of major input and output is assumed to be at the designed capacity factor (NETL, 2009) 16

17 Appendix B: Evaluation of cash flow We estimate the cash flow, using two concepts of assets valuation, free cash flow to firm (FCFF) and free cash flow to equity (FCFE) based on generally accepted accounting principles. The FCFF is the prior to debt cash flow to pay out to the firm s investors. It is expressed in the form of earnings before interest and taxes (EBIT) multiplied by one minus tax rate at 38% with adding depreciation but subtracting the capital expenditures and netting out the changes of working capital (ΔWC), as in Mun (2006): FCFF = EBIT*(1-Tax Rate) + Depreciation - Capital expenditures ΔWC In order to outline the financing debt structure we also estimate free cash flow to equity that can be paid to the equity stockholders after meeting all expenses and debt repayment. FCFE expression starts from net income with subtraction capital expenditures and principal repayment on debt, and addition the depreciation and new debt proceeds, as in Damodaran (2006): FCFE = Net Income + Depreciation - Capital Expenditure ΔWC Principal Repayment + New Debt Proceeds The main difference between FCFF and FCFE is that FCFF is derived from EBIT, as revenues excluding financing capital flow, while FCFE is based on net income adjusting for interest and debt repayments. For an equity valuation, two types of debt with 20 years of repayment period include the senior debt (80% of total debt) and the subordinated debt (20%). The senior debt is loan guaranteed with 5,5% interest that is 228 basis points above the average annual 10-year constant-maturities Treasury rate in The subordinated debt interest rate is assumed to be 9%, LIBOR plus 5.5%. The start of repayment for each loan is the plant start-up year based on the assumption that the lenders grant the grace period during plant construction. The EBIT and the net income are estimated from difference between the operating revenue and the operating expenses and costs. The operating revenues and the operating expenses are determined by placing values on the amount of resources from the purchase of input necessary for the CBTL production and selling its output in the market. The total operating cost is the sum of the operating expenses and the O&M costs. The O&M costs comprise a variable and a fixed O&M cost, including labor cost, cost of consumables, maintenance and other costs. We estimate the O&M cost as a percentage of an engineering, procurement, and construction (EPC) cost, such as the variable cost 1.5% and the fixed cost 3% of EPC cost using NETL Power Systems Financial Model 6.6 (Worhach, 2011). The EPC cost is used from the NETL (2009) study as a lump sum amount with their additional increase by 20%. We assume that the EPC cost is spread evenly during construction period before the cost escalation. The depreciation is determined as a declining percentage of the total capital costs and financing charges through NETL Power Systems Financial Model 6.1 (Worhach, 2011). The change in working capital reflects the change in cash that is required for day-to-day operations. It is calculated in each year as the accounts receivable less accounts payable using the PSFM 6.1 (Worhach, 2011). Days payable and receivable are assumed to be 30 days. The principal repayment and the new debt proceeds are based on the debt financing structure, interests and other financing parameters. 17

18 Appendix C: DCF, sensitivity analysis and simulation results Figure C1: Total operating expenses and O&M costs by year Millions of dollars $900 $600 $300 $191.1 $195.0 $81.9 $83.6 $402.6 $ % 13% 58% $ Operating Expenses Variable O&M Cost, $M Fixed O&M Cost, $M Figure C2: Product revenue and free cash flow from the CBTL plant Millions of dollars Millions of dollars $3,000.0 $1,500 $2,500.0 Product revenue $1,000 Free cash flow to firm $2,000.0 $1,500.0 $1,000.0 $500 $- $(500) $(1,000) $500.0 $(1,500) $ Year $(2,000) Year 18

19 NPV Table C1: Static effects of parameters on the NPV8 from the FCFF for the CBTL plant Parameters NPV8 (FCFF) $767 millions of dollars Input-output changes output downside output upside effective range (-10%) (+10%) base value ULSD price, $/gallon $96.55 $ $ $2.1 $2.5 $2.3 Operating capacity $ $ $ Discount rate $ $ $ % 9% 8% Investment outlays, $ million $ $ $ $5,035.4 $6,154.4 $5,594.9 Tax rate $ $ $ Crude oil prices,$/bbl $ $ $ $71.5 $87.3 $79.4 Coal, $/t $ $ $ $40.1 $49.0 $44.6 Inflation $ $ $ % 2.2% 2.0% Fixed O&M cost, $ million $ $ $ $165.3 $202.1 $183.7 Construction, yrs $ $ $ Variable O&M cost, $ million $ $ $91.58 $70.9 $86.6 $78.7 Escalation $ $ $ % 3.96% 3.60% Biomass, $/t $ $ $60.00 $81.0 $99.0 $90.0 Slag, $/t $ $ $12.47 $14.6 $17.9 $16.2 On-stream factor, % $ $ $ % 99% 90% Depreciation, % $ $ $ % 4.9% 4.5% Chemicals, $/t $ $ $12.07 $900 $1,100 $1,000 Water, t/d $ $ $4.78 $0.2 $0.3 $0.3 Plant Life, years $ $ $ Sulfur, $/t $ $ $2.01 $9.0 $11.0 $10.0 Construction period, months $ $ $ Millions of dollars Figure C3: Effects of major parameters on the NPV8 from the FCFF $1,600 $1,400 $1,200 $1,000 $800 $600 $400 $200 $0 ULSD Price, $/g. Operat. Capacity Investment, $M. Tax Rate Crude Oil,$/bbl Coal, $/t Biomass, $/t Inflation Chemicals, $/t Fixed O&M Cost, $M Change, % -10% -5% 0% 5% 10% 19

20 Figure C4: Impacts of sensitive variables ( 10 %) on the NPV12 from the FCFE for the CBTL plant Millions of dollars 20

21 Figure C5: The fitted probability distributions for the selected sensitive variables Bituminous coal real prices ($/t), Imported crude oil real prices ($/bbl), Dollars Dollars Fitted Distribution Lognormal Mean Sigma P-Value 0.57 Fitted Distribution Lognormal 3 Mean Sigma P-Value 0.89 The refinery capacity utilisation (%), ULSD wholesale/resale prices (cents/galon), Jan.07-Feb.11 Percent Cents Fitted Distribution Gumbel Minimum Alpha Beta 4.89 Mean P-value 0.91 Fitted Distribution Lognormal3 Mean Sigma P-value 0.83 Source data: U.S. EIA, available at: 21

22 Figure C6: CBTL IRR for FCFF forecast after 4000 simulations Mean IRR 0.087; standard deviation 0.01 IRR Figure C7: CBTL NPV12 from FCFF and IRR on equity forecast after 4000 simulations Mean NPV -$1.05 million; st. dev. $487million Mean IRR 0.12%; st. dev. 0.02% Millions of dollars IRR 22