The primary aim of this report is to help negotiate a Power Purchase Agreement (PPA) for the proposed hybrid gas-Linear Frensel Reflector (LFR) plant at Collinsville, Queensland, Australia.  The report’s wider appeal is the discussion of the current situation in Australian National Electricity Market (NEM) and techniques and methods used to model the NEM’s demand and wholesale spot prices for the lifetime of the proposed plant.

Executive Summary

1        Introduction

This report primarily aims to provide both dispatch and wholesale spot price forecasts for the proposed hybrid gas-solar thermal plant at Collinsville, Queensland, Australia for its lifetime 2017-47.  These forecasts are to facilitate Power Purchase Agreement (PPA) negotiations and to evaluate the proposed dispatch profile in Table 3.  The solar thermal component of the plant uses Linear Fresnel Reflector (LFR) technology.  The proposed profile maintains a 30 MW dispatch during the weekdays by topping up the yield from the LFR by dispatch from the gas generator and imitates a baseload function currently provided by coal generators.  This report is the second of two reports and uses the findings of our first report on yield forecasting (Bell, Wild & Foster 2014b).

2        Literature review

The literature review discusses demand and supply forecasts, which we use to forecast wholesale spot prices with the Australian National Electricity Market (ANEM) model.

The review introduces the concept of gross demand to supplement the Australian Electricity Market Operator’s (AEMO) “total demand”.  This gross demand concept helps to explain the permanent transformation of the demand in the National Electricity Market (NEM) region and the recent demand over forecasting by the AEMO.  We also discuss factors causing the permanent transformation.  The review also discusses the implications of the irregular ENSO cycle for demand and its role in over forecasting demand.

Forecasting supply requires assimilating the information in the Electricity Statement of Opportunities (ESO) (AEMO 2013a, 2014c).  AEMO expects a reserve surplus across the NEM beyond 2023-24.  Compounding this reserve surplus, there is a continuing decline in manufacturing, which is freeing up supply capacity elsewhere in the NEM.  The combined effect of export LNG prices and declining total demand are hampering decisions to transform proposed gas generation investment into actual investment and hampering the role for gas as a bridging technology in the NEM.  The review also estimates expected lower and upper bounds for domestic gas prices to determine the sensitivity of the NEM’s wholesale spot prices and plant’s revenue to gas prices.

The largest proposed investment in the NEM is from wind generation but the low demand to wind speed correlation induces wholesale spot price volatility.  However, McKinsey Global Institute (MGI 2014) and Norris et al. (2014a) expect economically viable energy storage shortly beyond the planning horizon of the ESO in 2023-24.  We expect that this viability will not only defer investment in generation and transmission but also accelerate the growth in off-market produced and consumed electricity within the NEM region.

2.1     Research questions

The report has the following overarching research questions:

What is the expected dispatch of the proposed plant’s gas component given the plant’s dispatch profile and expected LFR yield?

What are the wholesale spots prices on the NEM given the plant’s dispatch profile?

The literature review refines the latter research question into five more specific research questions ready for the methodology:

  • What are the half-hourly wholesale spots prices for the plant’s lifetime without gas as a bridging technology?
    • Assuming a reference gas price of between $5.27/GJ to $7.19/GJ for base-load gas generation (depending upon nodal location;) and
    • for peak-load gas generation of between $6.59/GJ to $8.99/GJ; and
    • given the plant’s dispatch profile
  • What are the half-hourly wholesale spots prices for the plant’s lifetime with gas as a bridging technology?
    • Assuming some replacement of coal with gas generation
  • How sensitive are wholesale spot prices to higher gas prices?
    • Assuming high gas prices are between $7.79/GJ to $9.71/GJ for base-load gas generation (depending upon nodal location); and
    • for peak-load gas generation of between $9.74/GJ to $12.14/GJ; and
  • What is the plant’s revenue for the reference gas prices?
  • How sensitive is the plant’s revenue to gas as a bridging technology?
  • How sensitive is the plant’s revenue to the higher gas prices?
  • What is the levelised cost of energy for the proposed plant?

3        Methodology

In the methodology section, we discuss the following items:

  • dispatch forecasting for the proposed plant;
  • supply capacity for the years 2014-47 for the NEM;
  • demand forecasting using a Typical Meteorological Year (TMY); and
  • wholesale spot prices calculation using ANEM, supply capacity and total demand
  • define three scenarios to address the research questions:
    • reference gas prices;
    • gas as a bridging technology; and
    • high gas prices.

The TMY demand matches the solar thermal plant’s TMY yield forecast that we developed in our previous report (Bell, Wild & Foster 2014b).  Together, these forecasts help address the research questions.

4        Results

In the results section we will present the findings for each research question, including

  • the TMY yield for the LFR and the dispatch of the gas generator given the proposed dispatch profile in Table 3;
  • Average annual wholesale spot prices from 2017 to 2047 for the plant’s node for:
    • Reference gas prices scenario from $18/MWh to $38/MWh
    • Gas as a bridging technology scenario from $18/MWh to $110/MWh
    • High gas price scenario from $20/MWh to $41/MWh
  • The combined plants revenue without subsidy given the proposed profile:
    • Reference gas price scenario $36 million
    • Gas as a bridging technology scenario $52 million
    • High gas price scenario $47 million

5        Discussion

In the discussion section, we analyse:

  • reasons for the changes in the average annual spot prices for the three scenarios; and
  • the frequency that the half-hourly spot price exceeds the Short Run Marginal Cost (SRMC) of the gas generator for the three scenarios for:
    • day of the week;
    • month of the year; and
    • time of the day.

If the wholesale spot price exceeds the SRMC, dispatch from the gas plant contributes towards profits.  Otherwise, the dispatch contributes towards a loss.  We find that for both reference and high gas price scenarios the proposed profile in Table 3 captures exceedances for the day of the week and the time of the day but causes the plant to run at a loss for several months of the year.  Figure 14 shows that the proposed profile captures the exceedance by hour of the day and Figure 16 shows that only operating the gas component Monday to Friday is well justified.  However, Figure 15 shows that operating the gas plant in April, May, September and October is contributing toward a loss.  Months either side of these four months have a marginal number of exceedances.  In the unlikely case of gas as a bridging scenario, extending the proposed profile to include the weekend and operating from 6 am to midnight would contribute to profits.

We offer an alternative strategy to the proposed profile because the proposed profile in the most likely scenarios proves loss making when considering the gas component’s operation throughout the year.  The gas-LFR plant imitating the based-load role of a coal generator takes advantage of the strengths of the gas and LFR component, that is, the flexibility of gas to compensate for the LFR’s intermittency, and utilising the LFR’s low SRMC.  However, the high SRMC of the gas component in a baseload role loses the flexibility to respond to market conditions and contributes to loss instead of profit and to CO2 production during periods of low demand.

The alternative profile retains the advantages of the proposed profile but allows the gas component freedom to exploit market conditions.  Figure 17 introduces the perfect day’s yield profile calculated from the maximum hourly yield from the years 2007-13.  The gas generator tops up the actual LFR yield to the perfect day’s yield profile to cover LFR intermittency.  The residual capacity of the gas generator is free to meet demand when spot market prices exceed SRMC and price spikes during Value-of-Lost-Load (VOLL) events.  The flexibility of the gas component may prove more advantageous as the penetration of intermittent renewable energy increases.

6        Conclusion

We find that the proposed plant is a useful addition to the NEM but the proposed profile is unsuitable because the gas component is loss making for four months of the year and producing CO2 during periods of low demand.  We recommend further research using the alternative perfect day’s yield profile.

7        Further Research

We discuss further research compiled from recommendation elsewhere in the report.

8        Appendix A Australian National Electricity Market Model Network

This appendix provides diagrams of the generation and load serving entity nodes and the transmission lines that the ANEM model uses.  There are 52 nodes and 68 transmission lines, which make the ANEM model realistic.  In comparison, many other models of the NEM are highly aggregated.

9        Appendix B Australian National Electricity Market Model

This appendix describes the ANEM model in detail and provides additional information on the assumptions made about the change in the generation fleet in the NEM during the lifetime of the proposed plant.

Publication Details