Review of Bulk Energy Storage in SEM

1. Mullany Engineering Consultancy Review of Bulk Energy Storage in SEM Gaelectric Energy StorageT&SC Modification Working Group 6th September 2012 Energy, from Utility to User

2. Working Group TOR • Define “Technical characteristics & capabilities of a CAES unit, and of other units having similar characteristics ” • Proposer: “impactsof registering in the SEM and the Market Rules as currently drafted” • Whether one or more existing registration options could be amended to allow registration of a CAES unit (perhaps by changing definitions in the TSC). • Examine the current Special Unit clauses to see if they can accommodate such units • Identify potential changes to the TSC • What are the likely high level impacts (time, cost and resources to implement option)? • How the identified option(s) further(s) the objective of the Demand Side Vision and other SEM • Deliver a recommendation to SEM Committee with a drafted modification Mullany Engineering Consultancy

3. Evaluation Criteria Impacts (TOR) • Assess the options in terms of the following questions: • The Working Group should consider how the following items, and others that may be discussed, can be addressed with respect to the option by which registration of CAES Units and other energy Storage Units in the TSC will occur. • Offering & Scheduling • Commercial Offer data • Technical Offer Data • Other data provision & sharing • Energy Settlement • Calculation of Eligible Availability • Calculation of Capacity Payments • When in generation mode • When in pumping mode • Calculation of Constraint Payments and Charges • Calculation of Uninstructed Imbalance Payments and Charges • Credit & Settlement • Inclusion in the MSP software and uplift • Treatment of Energy Storage Units when under test • Committee policy objectives? Mullany Engineering Consultancy

4. Policy Objectives • We also need to agree as a group what a unit registration in SEM should facilitate for that unit • While each proposal must understand how it impacts the list of the TOR evaluation criteria, e.g. Energy Settlement, these more flow from wider SEM policy • The policy must be the first high-level port of call when evaluating options, not the detail • Although the detail is important and ultimately required Mullany Engineering Consultancy

5. Our Objectives for this Meeting • Long-term: a method for registering storage technologies in the SEM that allows them to recover their short-run costs and compete equitably within the market schedule • Today: • A restatement of SEM Policy, becoming the first set of agreed evaluation criteria for our discussions • Evaluation of existing SEM options against those criteria • Agreement that existing options are not sufficient • Listing/workshop alternative structures, agreeing next steps Mullany Engineering Consultancy

6. CAES Overview Compression Train Generation Train Well Head 134MW generation 27MW minimum 81MW load ~53 MW minimum Depth • At stated figures: • 8 hours running • 10 hours to fill Storage Cavern Δp pmax pmin Mullany Engineering Consultancy

7. CAES and Other Storage • CAES running costs comprise: • The electrical energy consumed to compress air • Avoidable import UoS when consuming power * • The gas consumed in the running of the generator * • Avoidable O&M * • We believe what is missing from the T&SC is the ability to ensure those costs marked *are recoverable • We believe recovery of such short run energy costs is important for all storage technologies • A project needs to evaluate these SEM revenues alongside other DS3 revenues Mullany Engineering Consultancy

8. SEM Policy Evaluation Criteria • Apart from the market structures allowing the recovery of appropriate costs, insofar as possible we should have… • Market schedule should reflect generator characteristics to lower constraint costs for consumers • System operator should have correct data to minimise cost of production in SO dispatch • Commercial offer structures should allow transparent compliance demonstrable to RAs under BCOP • Market revenues earned by generators are independent of SO dispatch, based on the context of the COD and TOD Mullany Engineering Consultancy

9. Insofar as possible… • Market does not always succeed in getting COD and TOD to match generator capabilities • Increasing price quantity pairs for CCGT • Generator mode and fuel switching • These are limitations caused by the technical / time feasibility of solving the MSP software, and therefore not “unduly discriminatory” • This does not mean, however, that a “principle of acceptance” has been established that generators should be OK with managing unreflective cost structures Mullany Engineering Consultancy

10. Existing Options and Underlying Assumptions • Option 1: “The Split Unit” • Demand Side Unit plus Energy Limited Generator Unit • Potentially “negative generator” instead of DSU • Option 2: Pumped Storage • With adjusted efficiency to capture running costs • Our Assumptions • Generators, “negative generators” and pumped storage demand subject to full central dispatch • Demand Side Units subject to demand reduction dispatch from a nominated schedule Mullany Engineering Consultancy

11. Option 1: The Split Unit Demand Side Unit Energy Limited Generator Unit Submits Running Schedule to Market, COD, TOD Submits COD and TOD to Market Dispatched down by System Operator Dispatched by System Operator EnergyLimit Actual Availability, Dispatch, Metered Generation Ex Post Actual Availability, Dispatch, Metered Generation Ex Post MSQ Calculated, Constraint Payments MSQ Calculated, Constraint Payments Dispatch of the demand side impacts the energy limit of the Generator Unit Mullany Engineering Consultancy

12. Option 1: The Split Unit Challenges • Commercial Offer Data • Both generation and demand have to forecast SMP in making their offer • DSU must forecast a) low demand price and b) when it occurs in nomination • Generation offer made up of cost of demand and overall running costs • Storage has to forecast market prices making BCOP difficult to comply with • Energy Limit • Energy limited plant is required to correct energy limits in TOD to minimise cost of constraints • Storage is reliant on SO DSU dispatch to determine stored energy • Generator MSQ revenues are dependent on stored energy • Revenues are not independent of SO dispatch Mullany Engineering Consultancy

13. Option 1: The Split Unit Challenges • Restrictive offering of service • Unit owner submitting demand profiles • Demand dispatch down only… • Offer structure limits capability of machine to reduce cost of production A Demand Side Unit service is an overly restrictive operation of a storage device, and is not supported by Gaelectric.(This also includes making the demand not centrally dispatchable via retail registration in a standard Supplier Unit) Mullany Engineering Consultancy

14. Option 1: Further Commentary on DSU / Standard Supplier Unit • Power purchased like any other demand customer • This assumes that storage would have the right to manage its own demand schedule • When storing, demand has direct impact on generator availability – implication is generator would be self-decommitting • Alternatively, demand is fully dispatchable and outside of market • Penal relative to existing precedent of pumped storage rules; charged demand price Mullany Engineering Consultancy

15. Option 1: Split Unit with Negative Generator… • Will improve offered flexibility to system operator • But may need to have some form of maximum stored energy limit to prevent non-technical air compression… • Issues with BCOP compliance remain • Any structure that relies on storage submitting a definite price will not be transparent • Dependence of revenue on SO dispatch likely to become more difficult • SOs may determine based on COD not to store any energy… Mullany Engineering Consultancy

16. Option 2: Pumped Storage • Standard pumped storage unit with efficiency increased to capture running costs • What does this mean? See some example on next slide… • Note that efficiency increase needs to assume a relationship between forecast SMP and the running costs • Again, this may pose issues with BCOP • But let’s continue to examine this option to see how it evaluates against the other criteria • Constraint costs • Cost of production • Revenue dispatch independent • Generator recovering revenues Mullany Engineering Consultancy

17. Option 2: Standard Efficiency (Standard Storage Technology) Efficiency = 0.7MWh. 1MWh storage. Running costs to be recovered = €30 SMP = €20/MWh SMP = €60/MWh MSQ Gen 0.7MWh 1MWh Time (one Trading Day) Pump Revenue = -€20 + €42 = €22; Have not recovered the running costs Mullany Engineering Consultancy

18. Option 2: Adjusted Efficiency (Standard Storage Technology) Efficiency at 1.2. We assume that plant due to energy limit is not impacting prices SMP = €20/MWh SMP = €60/MWh MSQ Gen 1.2MWh 1MWh Time (one Trading Day) Pump Revenue = -€20 + €72 = €52; (IR in previous example plus running costs) Mullany Engineering Consultancy

19. Option 2: Standard Efficiency (CAES) Efficiency = 1.3MWh (burning gas for CAES adds to energy release beyond that stored). 1MWh storage. Running costs to be recovered = €30. Lower high price for demonstration. SMP = €30/MWh SMP = €40/MWh MSQ Gen 1.3MWh 1MWh Time (one Trading Day) Pump Revenue = -€30 + €52 = €22; Have not recovered the running costs with “IR” Mullany Engineering Consultancy

20. Option 2: Adjusted Efficiency (CAES) Efficiency at 1.8. We assume that plant due to energy limit is not impacting prices SMP = €20/MWh SMP = €40/MWh MSQ Gen 1.8MWh 1MWh Time (one Trading Day) Pump Revenue = -€20 + €72 = €52; (IR in previous example plus running costs) Mullany Engineering Consultancy

21. Option 2: Adjusted Efficiency (CAES) Flat SEM Price Example… SMP = €20/MWh SMP = €20/MWh MSQ Gen 1.8MWh 1MWh Time (one Trading Day) Pump Revenue = -€20 + €36 = €16; Market will schedule plant with flat SMP Mullany Engineering Consultancy

22. Option 2: Pumped Storage Review • Constraint costs: We are deliberately enshrining technical inaccuracies into MSQ to manage commercial cost recovery • MSQ becomes deliberately infeasible to follow in physical dispatch – we enshrine increased constraint costs to the consumer into the code • And unfair MSQ pressure on other generators! • Cost of production: System operators are provided with inaccurate data to dispatch the system; • SO requirement to minimise production costs undermined • SMP revenue is dispatch independent: good! • But see constraint revenues below… • Constraint payments: They are not paid to generator • The lack of constraint payments to pumped storage means that generator is not guaranteed to recover costs Mullany Engineering Consultancy

23. Review of Existing Options • Split Unit: • Struggles with BCOP compliance • Revenue not dispatch independent • Generator not guaranteed to recover its costs (worse with negative generator) • Restrictive offer reduces value in reduction of production costs (for DSU) • Pumped Storage • BCOP compliance again difficult • Market revenue dispatch independent: good • Generator not guaranteed to recover its costs under constraint scenarios • Enshrining increased constraint costs on consumers • Makes minimisation of cost of production difficult with non-technically accurate offers Mullany Engineering Consultancy

24. Gaelectric Conclusions • No existing structure is fit for purpose • Gaelectric believes that potential viable options are: • “Interconnector Unit” style unit, with • similar flexibility to manage COD as IU, e.g. participant is allowed to manage offer to procure “storage” • similar to IU procuring BETTA energy • Pumped Storage Unit, with • explicit commercial costs considered in MSQ schedule • Constraint payments • Seek agreement on first statement, and commitment to work towards development of listed and further potential options… Mullany Engineering Consultancy