4.0 ERCOT Market Analysis

4.1 Methodology

Tenaska has combined the strengths of two well-known electric pricing models with its own proprietary data to create a powerful “toolkit” for forecasting electric market pricing. Major components of Tenaska’s forecasting methodology are as follows:

• The Aurora^XMP model forecasts long-term regional electricity prices;
• The PowerWorld™ model checks intra-regional price estimates and provides nodal price forecasts in the ERCOT market; and
• Tenaska’s internal expertise customizes available project data

4.1.1 The Aurora^XMP Model

The Aurora^XMP model, developed by EPIS, Inc. (EPIS), forecasts energy prices based on the marginal cost of dispatch within the existing resource stack.

As shown in Figure 4.1.1, Aurora^XMP inputs include new and existing generator characteristics, retirements, expansion unit characteristics, fuel price projections, demand projections and transmission limitations. Aurora’s™ outputs include hourly load and generator costs, which are fed into the PowerWorld™ model, and plant-level information on generation, revenues and production costs.

Figure 4.1.1 – Aurora Inputs and Outputs

4.1.2 The PowerWorld™ Model

The PowerWorld™ model takes the hourly load and generator costs provided by the Aurora^XMP model and re-dispatches electric generators to resolve transmission constraints. PowerWorld™ will forecast nodal energy prices and plant revenues based on this re-dispatched system.

PowerWorld is an extremely visual, high-voltage power system simulation and analysis package. The core of the software is a powerful solution engine, capable of efficiently performing power flow analysis on systems containing up to 100,000 buses.

PowerWorld provides the ability to optimally dispatch the generation in an area or group of areas while simultaneously enforcing transmission line and interface limits. PowerWorld can then calculate the marginal price to supply electricity to a bus, while taking into account transmission system congestion.

4.1.3 Proprietary Tenaska Data

Tenaska has used its extensive experience in the electric and gas markets to develop specific, accurate information not available to the general public. Quantitative proprietary data is incorporated into both the Aurora^XMP and PowerWorld™ models, yielding more accurate results. In addition, Tenaska’s electric transmission and marketing experts have unique insights that assist in interpreting the data produced by both models, providing more useful conclusions. Tenaska’s internal natural gas and energy demand forecasts have been replaced in this analysis the USA Department of Energy (DOE) Energy Information Administration’s (EIA) Natural Gas Price forecast and ERCOT’s energy demand forecast. These are widely accepted third-party forecasts that are publicly available

4.2 ERCOT Market Considerations

4.2.1 ERCOT Nodal Implementation and Design

In an effort to improve reliability and improve wholesale market efficiency and transparency, ERCOT launched its nodal market on December 1, 2010. The nodal market replaced the zonal market that ERCOT had used since 2001. ERCOT’s system is the first to use a common information network that allows utilities to interactively view and report changes in grid operations. ERCOT states that the “enhancements to overall market efficiency should translate into substantial savings for consumers.” A Public Utility Commission of Texas (PUCT)–commissioned analysis by an independent consultant estimated the nodal system will save consumers USD$5.6 billion over the first 10 years. According to an ERCOT press release these savings will realized by:

• Improved use of generation resources through unit-specific dispatch – selecting individual units based on lowest price rather than on generation portfolios;
• More efficient management of transmission congestion through market-based mechanisms;
• More accurate price signals that better indicate where new generation and transmission is most needed (and where it is not) for managing congestion and maintaining reliability; and
• Improved ability to efficiently and reliably integrate the increasing quantities of intermittent resources, such as wind and solar generating facilities.

In the prior zonal market design, ERCOT had four price zones and energy schedules grouped in portfolios, rather than by individual unit. The nodal market captures prices at more than 8,000 “nodes” or any point where energy is added or taken out of the grid, including transmission lines, generators, electrical busses, breakers, switches and other similar devices defined in the network model.²

4.2.2 Texas Renewable Portfolio Standard

In 1999 Texas adopted rules for a Renewable Energy Mandate. The Public Utility Commission of Texas (PUCT) established a renewable portfolio standard (RPS). The RPS was then updated in 2005. The RPS established a renewable-energy mandate of 5,880 MW by 2015. This mandate has already been exceeded. The state of Texas currently has more than 9 GW of operating wind capacity.

4.2.3 Competitive Renewable Energy Zones

In 2008, the PUCT assigned USD$4.93 billion of Competitive Renewable Energy Zone (CREZ) transmission projects to be constructed. The CREZ projects are designed to transmit 18,456 megawatts (MW) of wind power from West Texas and the Panhandle to highly populated metropolitan areas in Texas.³ The CREZ transmission projects are a key component to the forecasted power prices in the study. Currently, wind capacity in West Texas is constrained from moving to population centers in the central, eastern and southern parts of the state. This has caused an oversupply of power in the ERCOT West Zone and has resulted in lower prices in that zone. The CREZ transmission projects will allow wind capacity in the West to flow east to areas with higher prices. This will help alleviate the oversupply situation in the ERCOT West Zone, and thus will have a positive impact on prices in the ERCOT West Zone, which is where the Trailblazer project is located. Figure 4.2.3 is from the PUCT. It illustrates the CREZ projects that are expected to be completed.⁴

Figure 4.2.3 – CREZ Map

4.3 ERCOT Market Assessment Assumptions

4.3.1 Supply

4.3.1.1 Existing Resources

Tenaska began its analysis of existing resources by reviewing information available from Ventyx, a leading source of energy information. Tenaska then supplemented the Ventyx data with its own knowledge of the ERCOT system, to arrive at its list of existing resources. Table 4.3.1.1 shows the existing resources modeled in the analysis, broken down by fuel type:

Table 4.3.1.1 – Existing ERCOT Resources

* Wind capacity is de-rated to reflect availability at peak demand

Figure 4.3.1.1 shows a graphical representation of this data.

Figure 4.3.3.1 – Existing ERCOT Resources

4.3.1.2 Expected Capacity Additions

Known Additions

Tenaska began its projection of capacity expansions by identifying capacity additions that currently are under construction. Tenaska considers such projects to have an extremely high probability of actually reaching commercial operation. Announced projects not yet under construction were not considered in the analysis, as it is extremely difficult to determine which announced projects actually will be built. Using only capacity additions under construction provides a defendable view of the market. Tenaska also obtained information on projects currently under construction from Ventyx and supplemented the Ventyx data with its own knowledge of the ERCOT systems. Table 4.3.1.2 shows the known capacity additions.

Table 4.3.1.2 – Known ERCOT Capacity Additions

Projected Generic Capacity Additions

Tenaska used the long-term system expansion capability within Aurora^XMP to determine

the optimal mix of new resources to be added to the market over time so that market reserve margins do not drop below a minimum reserve margin target. ERCOT currently has a minimum reserve margin target of 13.75%. Aurora^XMP chooses from new resource alternatives based on the net present value of hourly market values. Aurora^XMP compares those values to existing resources in an iterative process to optimize the set of new units.

Tenaska used proprietary input characteristics of potential new generation additions, including unit capacities and heat rates, variable O&M, fixed O&M, and operating characteristics related to the model’s commitment logic. Technologies considered include natural gas-fueled combined cycle and peaking units, coal (with CO₂ capture), nuclear, solar, and wind. The model was constrained so that it did not consider the addition of a traditional pulverized coal facility without carbon capture, as Tenaska believes it is unlikely that future coal plants will be constructed in Texas without CO₂ capture. Given the current lack of impetus in the USA toward cap and trade or other pricing mechanisms, the analysis assumes no price on carbon emissions.

Through 2030, Aurora^XMP selected only natural gas-fired unit additions, largely because of the low natural gas prices projected during that period. Lower natural gas prices only serve to widen Trailblazer’s commercial gap, and highlight the need for government policies that recognize the value of building baseload resources and capturing CO₂.

The results of this analysis are shown graphically in Figure 4.3.1.2.

Figure 4.3.1.2 – Projected ERCOT Capacity Additions

Retirements

The analysis takes into account the very small number of announced retirements in ERCOT. It should be noted that MWs to be retired represent an extremely small percentage of the total ERCOT supply. They do not affect the results of the analysis in a meaningful way because their high dispatch costs result in little or no actual dispatch into the market.

4.3.2 Demand

For annual demand growth, Tenaska used the2010 ERCOT Planning Long-Term Hourly Peak Demand and Energy Forecast. The monthly and hourly demand shapes were taken from EPIS. They are based on historical data from Federal Energy Regulatory Commission (FERC) Form 714, EIA 411 filings for demand and energy as well as reported hourly load from independent system operators. Figure 4.3.2.1 shows the projected ERCOT peak demand for each year of the study period. Figure 4.3.2.2 illustrates the hourly and monthly shape applied to the annual demand forecast for the year 2015.

Figure 4.3.2.1 – Forecasted ERCOT Peak Demand

Figure 4.3.2.2 – ERCOT Hourly Demand Shape Example, Year 2015

4.3.3 Emissions Pricing

Emissions price forecasts from Wood Mackenzie and CERA have been used as inputs in the model for both sulfur dioxide (SO₂) and nitrogen oxide (NO_x) emissions. Both third party forecasts are well respected and provide insight into emissions prices beyond the traded forward market. There is no price for CO₂ emissions assumed in the forecast. This is due to the fact that proposed Federal action on a cap and trade program for CO₂ has stalled, which has left uncertainty in the market over what the cost of CO₂ will be for generators in the future. It would be expected that any price on CO₂ would directly increase prices in ERCOT. This would directly increase energy margins for the Project, because it will be able to capture almost all of its CO₂ while marginal gas/coal units that set the price in ERCOT do not have CO₂ capture and have greater CO₂ emission rates.

4.3.4 Projected Fuel Prices

As discussed above, one of the key drivers for the valuation of the Project is the fuel price forecast. Tenaska used the EIA’s 2011 Annual Energy Outlook Early Release for the natural gas price forecast. Tenaska applied its proprietary gas basis assumptions to the EIA forecast to develop the regional gas price. Tenaska also uses plant-level coal price forecasts provided by Wood Mackenzie to accurately model the delivered cost of coal to all coal units in ERCOT. Wood Mackenzie’s coal price forecast is not made publically available so Figure 4.3.4.1 shows the gas and coal pricing used by the EIA to show the relative fuel price difference as projected by the EIA.

Figure 4.3.4.1 – EIA Fuel Price Forecast (USD)

Figure 4.3.4.2 incorporates the assumed market fundamentals and shows where the Project ranks from a dispatch cost perspective relative to the rest of the generating facilities in ERCOT. As discussed in Section 1.3.1, an energy-only market like ERCOT incents units to bid based on their variable operating costs to ensure that the units get called to dispatch.

Figure 4.3.4.2 – Trailblazer Dispatch Cost Relative to other ERCOT Generation

Trailblazer is able to bid into the market at a lower cost than other generators due to the additional revenue it will receive from sales of CO₂ for EOR. This CO₂ revenue allows Trailblazer to offset the majority of its fuel and variable O&M costs. Average coal dispatch costs in the chart equal USD$24/MWh in the year 2015. So while Trailblazer’s capital costs make the Project uneconomic, as discussed in Section 7, its low fuel and variable O&M costs, along with the revenues it will receive from CO₂ sales, would allow it to bid in to the market at a price that would ensure that the Project would dispatch virtually all the time and receive the clearing price for the highest cost unit each day. The difference between Trailblazer’s dispatch costs and the market clearing price allows Trailblazer to recover some of its fixed costs.

4.4 Results

By simulating hourly supply and demand and incorporating each unit’s forecasted dispatch cost, the model is able to produce an hourly power price forecast for the study period that is incorporated into Trailblazer’s energy revenue for the hours it dispatches to the market. It should be noted again that for this report the model used publicly available information, rather than Tenaska’s proprietary information, for several key model inputs. Therefore, the prices shown in Table 4.4 below do not represent Tenaska’s view on the forecasted power prices in ERCOT. They do, however, provide an indicative indication of the pricing Trailblazer might see in the market.

Table 4.4 – Annual Forecasted Power Prices (USD) in the ERCOT West Zone