Building a resilient grid for power and gas

  • Slides: 42
Download presentation
Building a resilient grid for power and gas markets Nicola Falcon Group Manager, Forecasting

Building a resilient grid for power and gas markets Nicola Falcon Group Manager, Forecasting

About AEMO We operate Australia's National Electricity Market and power grid in Australia’s eastern

About AEMO We operate Australia's National Electricity Market and power grid in Australia’s eastern and south-eastern seaboard, and the Wholesale Electricity Market and power grid in south-west WA. Both markets supply more than 220 terawatt hours of electricity each year. We also operate retail and wholesale gas markets across south-eastern Australia and Victoria’s gas pipeline grid. Collectively traded more than A$20 billion in the last financial year. Ownership 40% 60% Governments Market participants of Australia

Our industry is in disruption

Our industry is in disruption

FROM Here’s what’s Energy mix changing and infrastructur e are transforming TO A static

FROM Here’s what’s Energy mix changing and infrastructur e are transforming TO A static world: Rapidly changing world: Generation q Predictable demand growth and demand mixon coal q Predominantly based q Consumption flat, but demand peaks even more pronounced under extremes and gas resources q A power system designed around bulk energy transport on main highways from major (synchronous) generation centres q Supplies involve geographically dispersed, technologically diverse resources q. Weather Requiring: Falling impacting q flexible dispatchablecosts plant of demand renewables q energy storage q visibility and controllability of resources, including embedded q Efficient re-configuration of the transmission system to support

Drivers of disruption Diverse, Variable, Unpredictable Supply and Demand Weather and climate changes Supply

Drivers of disruption Diverse, Variable, Unpredictable Supply and Demand Weather and climate changes Supply sources Reduced Resiliency Ageing infrastructure Faster and More Granular New Areas of Vulnerability Energy industry New service models Unknown Policy uncertainty Reduced Visibility Cyber security Consumer preferences Electronic vs Synchronised Resources

Here’s what’s changing Changes in resources In 2013 there were 22 active projects totaling

Here’s what’s changing Changes in resources In 2013 there were 22 active projects totaling 1, 231 megawatts Generation and demand mix New Peak demand Fast forward to 2018, there are currently over 136 connection requests totaling 19, 507 megawatts South Wales 14, 700 MW Current capacity Weather impacting 18, 900 MW demand New connections 47, 000 MW Coal retirements 1, 680 MW in 2022 Falling costs of renewables

More capacity required to deliver demand Here’s what’s changing Projected changes in scale of

More capacity required to deliver demand Here’s what’s changing Projected changes in scale of resources Generation and demand mix Weather impacting demand Falling costs of renewables

The rapid solar and storage consumer uptake Here’s what’s Changes in changing customer behaviour

The rapid solar and storage consumer uptake Here’s what’s Changes in changing customer behaviour A solar panel is being installed every 6. 5 minutes in Australia Generation and demand mix Weather impacting demand Falling costs of renewables

Here’s what’s Australia onchanging track to be the world ’s most decentralised electricity system

Here’s what’s Australia onchanging track to be the world ’s most decentralised electricity system Generation and demand mix Weather impacting demand Falling costs of renewables

…contributing to a flat demand outlook Here’s what’s changing Changes in customer behaviour Generation

…contributing to a flat demand outlook Here’s what’s changing Changes in customer behaviour Generation and demand mix Weather impacting demand Falling costs of renewables

Increased variability and flexibility 23 July 2017 in South Australia Here’s what’s changing Changes

Increased variability and flexibility 23 July 2017 in South Australia Here’s what’s changing Changes leading to operational challenges Generation and demand mix Weather impacting demand Falling costs of renewables

Needing a real focus on frequency and strength Fast Frequency Response Fault current leading

Needing a real focus on frequency and strength Fast Frequency Response Fault current leading to operational challenges Voltage phase angle Phase angle Here’s what’s changing Changes Generation and demand mix Weak System Strong System Weather impacting demand Falling costs of renewables

…and improved forecasting of future risks Here’s what’s changing Changes leading to reliability challenges

…and improved forecasting of future risks Here’s what’s changing Changes leading to reliability challenges Generation and demand mix zzzzz Weather impacting demand Falling costs of renewables

Here’s what’s changing Climate change Generation and demand mix Weather impacting demand Bushfires in

Here’s what’s changing Climate change Generation and demand mix Weather impacting demand Bushfires in winter Falling costs of renewables

Real concerns around gas shortages in next 5 yrs Here’s what’s changing Reduction in

Real concerns around gas shortages in next 5 yrs Here’s what’s changing Reduction in gas reserve estimates Generation and demand mix zzzzz Weather impacting demand Falling costs of renewables …leading to talk of LNG import terminals

Here’s what’s changing Cyber security Generation and demand mix Weather impacting demand Falling costs

Here’s what’s changing Cyber security Generation and demand mix Weather impacting demand Falling costs of renewables

Here’s what’s changing No worries! Generation and demand mix Weather impacting demand Falling costs

Here’s what’s changing No worries! Generation and demand mix Weather impacting demand Falling costs of renewables

The imperative to adopt 4 th industrial thinking 1 st 2 nd Mechanisation, water

The imperative to adopt 4 th industrial thinking 1 st 2 nd Mechanisation, water power, steam power Mass production, assembly line, electricity 3 rd 4 th Computer and automation Fusion of physical, digital and biological systems

Big Data and Analytics Characteristics of industry 4. 0 Internet of Things (Io. T)

Big Data and Analytics Characteristics of industry 4. 0 Internet of Things (Io. T) • Dynamic Simulation • Integrated systems • People systems Cybersecurity Industry 4. 0 • Extreme pace of change • Interconnected economies Cloud Computing Horizontal & Vertical Integration Additive Manufacturing Augmented Reality Robotics

Applying 4 th revolution approaches in Forecasting Adopting a system of systems approach Regulatory

Applying 4 th revolution approaches in Forecasting Adopting a system of systems approach Regulatory System Power System Business Systems Investments Systems Market System People Systems Data System

Power system Wind integration Distribution • Huge computational models forecasting • Multiple data inputs

Power system Wind integration Distribution • Huge computational models forecasting • Multiple data inputs from more sources • Digitalisation allows value to be determined at more granular level • Optimisation of entire supply chain PV integration Power Plant Demand side participation Transmission Active network management Electricity storage EV charging

Pre 2010 Data systems Applying AI to systems solutions for situational awareness 6 data

Pre 2010 Data systems Applying AI to systems solutions for situational awareness 6 data points per customer meter read 2015 2021 9, 000 data points – for five minute reads Over 100, 000 data points – near real time reads Huge growth in big data – from 6 data points to over 100, 000 data points annually Harnessing digitalisation to make things work Reducing barriers to entry by investing in whole grid Using data to manage the system in an efficient way Leveraging technology and assets

Investment system Need to optimise capital investment Where investments can be made to leverage

Investment system Need to optimise capital investment Where investments can be made to leverage assets and resources, creating value for consumers.

Strategic partnerships will be crucial Collaborate and partner with range of stakeholders to leverage

Strategic partnerships will be crucial Collaborate and partner with range of stakeholders to leverage capabilities where synergies exist

Using PLEXOS within AEMO Forging ahead with strong relationships

Using PLEXOS within AEMO Forging ahead with strong relationships

Market modelling and Forecasting at AEMO A systems thinking approach System Planning • Projecting

Market modelling and Forecasting at AEMO A systems thinking approach System Planning • Projecting the development of the electricity system over the next 20 years. • Understanding the transmission development needed. Reliability Forecasting • Uses Plexos to undertake market modelling • Forecasts reliability risks over the next 10 years. Demand Forecasting • Forecast energy consumption and peak demand • Understanding of new technologies such as EV, ESS, etc.

Long-term modelling approach • Electricity and gas co-opt • Least-cost planning • LDC approach

Long-term modelling approach • Electricity and gas co-opt • Least-cost planning • LDC approach • One-step solve • Simplified hydro representation • Regional topology • One weather pattern Integrated Model Detailed Long. Term Model • Electricity optimization • Least-cost planning • Fitted chronology • Multi-step solve • Simplified hydro representation • Regional topology • One weather pattern • Electricity optimization • Nash-Cournot algorithm • Time-sequential • Detailed hydro representation • Monte Carlo for outages • Regional topology • Fully constrained • 8 weather patterns and two maximum demand targets Short-Term Model

Integrated modelling Conversion/processing technologies Primary fuels Coal Electricity Modelling End use Industrial Power Generation

Integrated modelling Conversion/processing technologies Primary fuels Coal Electricity Modelling End use Industrial Power Generation Residential Renewables Natural Gas, LPG, Coal Seam Gas Oil Other: Biomass, Nuclear, etc. Gas Modelling Commercial Gas Processing Facilities Transport Refineries Hydrogen Production

…and all trade-offs increase iterative planning Here’s what’s changing Simulation time leads to tradeoffs

…and all trade-offs increase iterative planning Here’s what’s changing Simulation time leads to tradeoffs Generation and demand mix Fitted chronology with daily steps Approximation of intra-regional constraints Weather Multi-step solve impacting Linearised decisions demand Simplified hydro representation Falling costs of renewables

AEMO’s Reliability Forecastin g Models ESOO • 10 -year forecast, usually done once per

AEMO’s Reliability Forecastin g Models ESOO • 10 -year forecast, usually done once per year • Informs the market of the need for new investment • Identifies whether AEMO needs to procure additional reserve MT PASA • Forecast of the next two years, published weekly • Assists in operational decision-making PLEXOS

ESOO The 2018 ESOO showed a heightened risk of USE in Victoria in summer

ESOO The 2018 ESOO showed a heightened risk of USE in Victoria in summer 2018 -19. On 25 January load was shed due to insufficient supply in Victoria.

Probabilistic Forecasting Monte Carlo simulations USE assessment Account for different weather patterns Subject to

Probabilistic Forecasting Monte Carlo simulations USE assessment Account for different weather patterns Subject to Reliability Standard

AEMO’s reliability model The models used in ESOO and MTPASA are very similar. •

AEMO’s reliability model The models used in ESOO and MTPASA are very similar. • The configuration within these models is actually relatively simple, particularly when compared with other AEMO Plexos models such as those used in the Integrated System Plan. • The models focus on physical system capabilities, with the key inputs being: • • Generator ratings Forced outage parameters Demand traces Renewable generation traces • The complexity in AEMO’s reliability models is due to: • Scale • Transmission constraint equations

Generator outages drives simulation scale 8 Weather Patterns 2 Max Demand Targets 100 Outage

Generator outages drives simulation scale 8 Weather Patterns 2 Max Demand Targets 100 Outage Patterns Both ESOO and MTPASA require 1, 600 simulations!

Transmission constraints • Thousands of constraints are used in reliability assessments. • Uses custom

Transmission constraints • Thousands of constraints are used in reliability assessments. • Uses custom dll assemblies developed by EE to incorporate the complex constraint equations used in the NEM. • Constraints are read from an Excel workbook. • Also involves a multi-pass approach to accounting for RHS calculation. Formulations can be very complex Example RHS: row_0 = 963 - 25. 8; row_1 = 26. 7 - 55; row_2 = 10. 48 * unitstep('Connection PointTARONG#1'. 'Energy generation target' in MW, 0); row_3 = 10. 48 * unitstep('Connection PointTARONG#2'. 'Energy generation target' in MW, 0); row_4 = 10. 48 * unitstep('Connection PointTARONG#3'. 'Energy generation target' in MW, 0); row_5 = 10. 48 * unitstep('Connection PointTARONG#4'. 'Energy generation target' in MW, 0); row_6 = 32. 25 * unitstep('Connection PointSWAN_E'. 'Energy generation target' in MW, 0); ………………. . row_225 = (row_222 + row_224) * nunitstep(row_213, 0); row_226 = (row_225); row_227 = (row_215 + row_226) * nunitstep(row_213, 0); row_228 = (row_227); row_229 = (row_222 + row_224) * unitstep(row_213, 0); row_230 = (row_229); row_231 = (row_228 + row_230); row_232 = (row_209 + row_210 + row_231)^0. 5; row_233 = 43. 8 * (row_232); row_234 = 95; row_235 = 0. 00001; row_236 = 378; = row_0 + row_1 + row_22 + row_90 + row_151 + row_202 + row_233 + row_234 + row_235 + row_236;

How AEMO deals with scale Significant developments over the past 12 months

How AEMO deals with scale Significant developments over the past 12 months

Challenges in increasing simulation scale • Until 2018, the simulations were of a scale

Challenges in increasing simulation scale • Until 2018, the simulations were of a scale that could be managed by simulating across a number of machines and using the traditional solution file approach. • New analysis techniques were required that needed more detailed results than were previously produced (just USE outcomes). • The key limitations AEMO found in trying to increase the scale of simulations were: • Managing simulations across many machines using Plexos Connect was challenging. • The time taken to produce solution files for simulations of this scale. • Difficulty in extracting all the required information from so many solution files.

Improvements – moving to cloud and Plexos Connect • Cloud resources to handle the

Improvements – moving to cloud and Plexos Connect • Cloud resources to handle the bulk of our modelling and data workload. • AEMO uses VM Scale Sets: • Up to 51 high spec identical virtual machines which are built from scratch and operational within 10 minutes. • Each VM is customised for our modelling requirements, including the installation of the Plexos client. • Ability to scale up or down the number of VMs to match our simulation requirements. • Plexos Connect is used to manage the deployment of the simulations. • There have been significant improvements in the stability and reliability of Plexos Connect over the past 12 months which have allowed AEMO to manage simulations of this scale.

Improvements – managing result output • The results of simulations at this scale are

Improvements – managing result output • The results of simulations at this scale are unable to be managed by the traditional solution file approach. The time taken to stitch the hundred simulations together with many result properties required over thousands of objects is not feasible. • AEMO has developed a number of approaches, all of which stop Plexos from stitching solution files together. • Recent developments have focused on bypassing Plexos reporting entirely using custom dll assemblies to extract and store simulated results on the Cloud.

Key benefits of developments so far Enhanced analytics • Storing more information allows us

Key benefits of developments so far Enhanced analytics • Storing more information allows us to extract key insights Accuracy Efficiency • Increased simulation scale provides greater precision and accuracy • The significant improvements to simulation management and results processing has increased simulation speed

Development for the future – increased automation and embedding PLEXOS in the cloud •

Development for the future – increased automation and embedding PLEXOS in the cloud • Custom software that extracts inputs from DBs and file storage to build a Plexos model • Improved orchestration to manage the scaling of cloud resources • Continued improvement of the solution loader to allow greater flexibility and aggregations Case loader Improved orchestration Solution loader

Any questions?

Any questions?