1 / 21

Version 1.0

EPOC Winter Workshop 2008: New Zealand Wind Integration Study Goran Strbac, Danny Pudjianto, Anser Shakoor, Manuel J Castro Imperial College London Guy Waipara, Grant Telfar Meridian Energy Limited Sep 2008. Version 1.0. Background & motivation for NZ wind study. Anecdotes are not data

naoko
Download Presentation

Version 1.0

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. EPOC Winter Workshop 2008:New Zealand Wind Integration StudyGoran Strbac, Danny Pudjianto, Anser Shakoor, Manuel J CastroImperial College LondonGuy Waipara, Grant TelfarMeridian Energy LimitedSep 2008 Version 1.0

  2. Background & motivation for NZ wind study Anecdotes are not data Assertion is not analysis Worldwide the rise of wind power has in the past and will continue into the future to present unique challenges to power system planning Motivation for NZ study was to inform and shape the debate in NZ to move beyond gut reaction and anecdote to develop methodologies and quantify – from first principals – what unique costs wind generation brings to the system Work first mooted in late 2005 and begun in earnest in May 2006 having first identified Prof Strbac as a expert in the integration of distribution generation into power systems. A desire to learn sometimes hard won lessons from other jurisdictions Explicit intent to focus on additional system operation costs due to wind NOT to focus on issues of market pricing, dynamic efficiency, … Place wind intermittency within the context of an imperfect power system 1

  3. Challenges of integrating wind generation Generation capacity adequacy How “reliable” is wind generation as a source? How much conventional capacity can it displace? What are the system integration capacity costs and benefits? Wind generation is primarily an energy source with limited ability to provide reliable generation capacity at times of peak demand. Real time system balancing What are the needs for flexibility and reserve? What are the costs? What is the role of storage, demand side participation and inter-connectors? Additional requirements for instantaneous and frequency keeping reserves. Additional requirements for scheduling reserve. Transmission network requirements How much new transmission capacity is required to efficiently transport wind power? System stability What is stability performance of the system with new forms of generation? Can this technology contribute to improving stability? Role of enabling technologies Can storage and responsive demand have a role in facilitating integration of wind generation? Are these solutions competitive? What are the drivers of value? What new tools are required to support system management with wind generation? Technical, commercial and regulatory framework Are the technical, commercial and regulatory arrangements appropriate for a system with significant contribution of wind? Are the Grid Codes and Standards appropriate? Are the arrangements for access to transmission networks appropriate? Are non-network solutions to network problems competitive? Does the market reward flexibility adequately? 2

  4. Future Generation Scenarios

  5. NZ Electricity System: Future Scenarios Three scenarios corresponding to future generation snapshots in 2010, 2020 and 2030 were constructed to represent increasing levels of wind penetration: High Wind 2010: 2,100GWh (4.5%) NI: 432 MW SI: 203 MW NZ: 634 MW Very High Wind 2020: 6,700GWh (12.5%) NI: 1,434 MW SI: 632 MW NZ: 2,066 MW Very High Wind 2030: 10,700GWh (17.7%) NI: 2,215 MW SI: 1,197 MW NZ: 3,412 MW Explicit intent to focus on system operation with and without wind at various points in time NOT to focus on issues of dynamic efficiency, etc: Of course not really that black and white in choice of scenario and implied counterfactual 3

  6. Wind Profiles 4 • Wind data profiles derived from: • 10 minute real wind readings – mostly metering towers at 40m • 11 actual or proposed sites • Continuous 2 year period (Jan2005-Dec2006) 2005 wind profiles for 2010 scenario Very High Wind 2030 High Wind 2010 Very High Wind 2020

  7. 5 Initial Electricity Generation and Demand Scenarios Key Assumptions: • Load & generation are defined at the “station gate” (including embedded) • Annual demand growth is approximately 700GWh (110MW at peak) • No decommissioning of existing assets (except New Plymouth PS) • ‘Status quo’ fuel prices in perpetuity: gas = $6.5-7.5/GJ, coal = $4/GJ, distillate = $25/GJ

  8. Capacity Value and Additional Capacity Cost of Wind Generation in the NZ Electricity System

  9. Capacity Assessment Objectives Assessment of ‘optimal’ overall generation capacity requirements in each future wind scenario Capacity credit evaluation of wind generation Evaluation of additional capacity cost of wind generation Sensitivity Studies: Wind forecasting errors Impact of hydro conditions (Dry/Average/Wet) Impact of inter-connector size and its reliability Impact of wind diversity 6 Wind-Hydro-Thermal System

  10. Capacity Assessment Formulation The system reliability criterion for capacity adequacy applied in this study is Loss of Load Expectation (LOLE) with a conservative target of 8 hours/year. Hydro is used to minimize overall thermal capacity requirements (ie the objective function) Hydro is aggregated at the island level (with key catchment level constraints maintained) Hydro is modelled as a fully reliable generation Weekly reservoir energy releases (from Spectra) are used – water cannot be shared between weeks Spare hydro capacity is available to contribute to system reliability Main constraints include: Aggregated (wind + hydro + thermal) production must meet demand in each time period Minimization of wind and hydro energy curtailment The aggregated reservoir size of each island for hydro energy storage Minimum reservoir levels (10% of reservoir size) 99% reliability of the DC inter-connector 7 2030: Without Wind 2030: With Wind

  11. System Capacity Margin Requirements & the Capacity Credit of Wind 8 Key assumptions: • Conservative system reliability criterion: LOLE < 8 hours/year • Availability of conventional generation: 85% (planning time horizons) • NI-SI Inter-connector reliability: 99% • Wind forecasting approach is conservative & accommodates 99% of the wind variations across 4 to 6 hours time horizon.

  12. Summary of Capacity Results Additional capacity costs attributed to wind generation Fall from 2010 to 2020 due to a larger HVDC size Rise by 2030 due to reserve requirements to accommodate larger wind forecasting errors Unlike peak focused thermal based power systems, the capacity value of wind in NZ is driven by its higher load factor as well as by large variations in a relatively small period of time Hydro increases capacity credit of wind however, but at higher penetration levels its contribution to firming up the wind power reduces Capacity credit of wind generation in the NZ’s hydro dominated system is higher than in the other thermal based systems, however, it also declines with rise in wind penetration level Capacity values for wind are not effected by hydro (dry) conditions although the overall capacity requirements increase with low availability of hydro energy The low production of wind for several days is found not to effect the capacity value of wind as this is compensated by the flexible hydro energy with presence of large hydro reservoirs 9 • Key Assumptions: • Capacity cost of (OCGT) conventional plant = 100 $/kW/yr to 150 $/kW/yr • The additional costs of wind are relative to a base load thermal plant

  13. Cost of Wind Power Driven Reserve in the NZ Electricity System

  14. Wind Reserve Requirement Objectives Quantify operating reserves needed to deal with wind intermittency and wind forecasting errors Instantaneous reserve (up to 30 mins) provided by synchronised generators Frequency keeping reserve (up to 1 hour) provided by synchronised generators Scheduling reserve (for 4-6 hours) provided by synchronised and standing generators Evaluate wind related impact of increased reserve requirement on system operating costs Fuel cost, generation start up cost and no-load cost Cost of interruptible load Sensitivity analysis Hydro conditions (Dry/Average/Wet) Increased wind power (scenario 2010,2020,2030) Inter-connector capacity Location of future wind power (North and Southland scenarios) Different wind profiles (year 2005 and 2006 data based) 10

  15. Reserve Requirement Methodology Additional reserves increase operating costs Requires more on-line capacity Use of low merit (expensive) generation Lower efficiency-part loaded plants Increased frequency of start-ups of generators Increased demand of Interruptible load (IL) Increased standing reserve Operating cost is determined using a MILP generation scheduling optimisation model for energy production and the allocation of reserves among synchronised units Objective Minimise the total system generation costs (including no load, start up cost and IL costs) Operational constraints Power balance constraints Generation constraints: Min stable generation, Power rating, Max Instantaneous Reserve, Ramp up/down constraints, Min up/down time constraints, Load factor constraints for CCGTs (minimum 75%) Hydro catchment power constraints: Daily ROR energy constraints, Weekly hydro inflows constraints, Reservoir constraints Reserve constraints: Min instantaneous and frequency keeping reserve provision for each island (no reserves sharing) Flow constraints on HVDC Cost of additional operating reserve attributed to wind is determined as the difference between the operating cost of the system with and without wind reserve component Cost of standing reserve is determined by calculating the expected energy of standing reserve that would be exercised 11 North Island South Island Amount of reserve requirements increases with increased wind output

  16. Reserve Requirements & Assumptions Key Assumptions: Additional operating reserves in each Island analysed separately, i.e. (no operating reserve sharing across the DC) Power transfers across the DC link are limited by MW only (with the max of NI to SI level set to 60% of SI to NI) All operating spinning reserves quantities for instantaneous reserve and frequency keeping are unable to meet demand requirements Operating reserve mainly provided by part loaded plant along with a contribution from demand side (interruptible load) Standing reserve is provided by off-line thermal plants which can synchronise and produce electricity quickly Levels of operating reserve required cover about 99.5% of all operating conditions. Reserve from synchronised plant are assessed at 30 min and 1 hour. Standing reserves deal with wind forecasting errors beyond 1 hour Reserve requirements are computed for half-hourly. Additional reserve requirements to deal with wind variability never exceed expected wind power output. Day/night system inertia impacts on operating reserve requirements are modelled. CCGTs assumed to operate with minimum load factor of 75% irrespective of hydro and wind conditions. 12 • Note: all units are expressed in MW • IR (Instantaneous Reserve) • FK (Frequency Keeping) Case1: dominated by standing reserve Case2: balanced allocation Case3: dominated by spinning reserve

  17. Summary of Reserve Requirements Results Additional reserves are needed to cover the unpredictability of wind power: Instantaneous reserve provided by synchronised generators Frequency keeping reserve to cover 1 h wind variability provided by synchronised reserve Scheduled reserve to cover 4-6 h wind variability provided by synchronised + standing reserve The quantity of wind reserve component increases with rise in wind penetration 13 • Provision of scheduling reserve up to 4-6 hour time horizon • For low wind penetration (2010), hydro will be the primary source • For high wind penetration (2020 & 2030), it is desirable to use flexible standing power plants • The cost of additional reserve to deal with forecasting error of wind for several scenarios have been quantified • For low penetration (4.9% in 2010) , it is around 0.19 $/MWh of wind energy • It increases to 2.42 $/MWh of wind energy in 2030 with high wind penetration (17.9%) • Hydro remains the primary source of synchronized reserve, but in the future a greater contribution from IL and other thermal plants will be required Case1: dominated by standing reserve Case2: balanced allocation Case3: dominated by spinning reserve

  18. Summary of Findings

  19. Summary of Key Results 14

  20. Other Issues

  21. Other Wind Related System Issues … Quality of long-run wind record: Single biggest impediment to systematic analysis of wind However good correlation (daily energy+) between synthetic (NIWA weather station) data and actual wind farm data Annual wind variation: Positive correlation between wind and hydro may increase need for “hydro-firming” plant Not investigated here – but likely order of magnitude … small?: A 30 year Spectra run with 8TWh of wind generates a total system cost of $25B Removing wind variation decreases market prices by $2/MWh (2%) and system costs by $63M (0.3%) 15 • Correlation between demand and wind: • Positive correlation between extreme peak demand and low wind conditions may reduce capacity credit of wind and increase need for additional ‘firm’ system MW 2006 Calendar Year

More Related