Workstream 2 Dispatchability and Variability Management Studies

1
Workstream 2Dispatchability and Variability
Management Studies

• Leslie Bryans SONI
• Paul Smith ESB National Grid
• (for the Working Group)

2
Background and key issuesLeslie Bryans
Study scope, programme etc.Paul Smith
3
Background and key issuesLeslie Bryans
Study scope, programme etc.Paul Smith
4
The issues

• Effect of non-wind renewables upon the ability to
  manage
• Variability and uncertainty management from wind
  and like technologies
• Effect of better forecasting and advanced wind
  management schemes
• Effect of generation mix and load flexibility on
  the overall picture

5
Types of generation

• Renewable and Embedded Generation with
  traditional certainty
• Availability
• Certainty
• Wind and like technologies with high uncertainty
  / variability

6
Non-wind issues

• Suppose there was ???1000MW ??? of non-wind
  embedded generation
• What plant would the embedded generation
  displace?
• Would embedded generation be fully dispatchable
  (a size issue)?
• Would it provide needed ancillary services?
• If so its just like traditional generation
  albeit in different locations Grid Code
  provisions are useful but do not apply to very
  small plant
• If not then we need to understand what this
  ???1000MW??? means for system management?

7
Key questions

• For generation dispatch scenarios
• How much embedded is likely as
• dispatchable
• partially dispatchable
• non-dispatchable
• Where?
• When?

8
Energy balance and unit commitment / dispatch
with wind like technologies

• What do we mean by wind variability and
  uncertainty?
• How can we assess the impacts?
• How to manage wind variability and uncertainty?

9
Wind Energy Variability
10
Variability analysed
96 confidence
Be careful available not Operating level
11
Variability geo-diversity one farm and
all-island (44 farms)
Improvement 29 Probably saturates With
penetration
Improvement 75 probably Improves further With
penetration
12
Wind Variability and forecast
Real prediction from Anemos
13
(No Transcript)
14
Natural Uncertainty Better local forecasting
Improvable Uncertainty Better general forecasting
Gain by forecasting
15
Power System Reserve

• Rescue flows
• Load prediction errors
• Wind prediction errors

16
(No Transcript)
17
Power System Reserve

• Very Rapid 5 sec
• spinning reserve
• Disconnectable load
• Rapid
• Energy storage
• Medium
• Standing OCGTs and like plant 10 min
• Progressively longer
• Hot, warm and cold thermal plant 24 hour

No option Some options All options
18
System Inertia

• Non-synchronous generators may not contribute to
  system inertia
• This could hamper the ability to deal with
  frequency transients
• Will this limit the penetration of renewables?

19
Background and key issuesLeslie Bryans
Study scope, programme etc.Paul Smith
20
Workstream 2(b)

• Detailed accurate modelling of commitment and
  dispatch taking account of uncertainties
• Probably 3 later years each with 5 renewables
  scenarios
• Renewables, especially wind, time series will be
  required
• Plant capacity, replacement, emissions and cost
  assumptions
• Demand management assumptions
• Wind forecasting accuracy assumptions

21
Basic Presumption

• Governments want to maximise contribution of
  renewables to policy objectives such as
• Reduction of greenhouse gas and other emissions
• Reduction of dependence on external primary
  energy sources
• Reduction of future cost uncertainty
• Local economic and social benefits
• Security of energy supplies
• Changes in renewable generation output, either
  predictable or unpredictable, must be covered by
  demand management schemes, generation or
  interconnection to UK.

22
Basic Study Objective

• Model future operation of system so as to
  estimate
• Future operation costs
• Future plant emissions
• For a number of renewables scenarios
• Identify and analyse measures to facilitate
  greater levels of renewables

23
A modelling challenge

• Modelling current system operation for production
  cost studies, constraints studies, ancillary
  services budgets etc. is a challenge, even though
  we know how the system is operated.
• For this study, we need to model system operation
  where we dont yet know how the system will be
  operated

24
Tasks

• Propose and discuss methodology
• Develop commitment and dispatch model and test on
  2005 case
• Construct renewable generation time series
• Carry out detailed dispatch studies for base
  scenarios, estimating costs, emissions etc.
• Investigate options to improve dispatch,
  integrate more renewables
• Prepare overall report

25
2005 model verification case

• Plant mix as now
• Load as recorded
• Renewables as installed (using scaled 2004 wind
  data which is available and coded)

26
Validation

• Model variability matches observed pattern
• Estimation of variability change with wind farm
  penetration
• Dispatch model mirrors real dispatches
• Models are extensible with similar error rates

27
Scenarios

• 2005 validate model
• 2010
• 2015
• 2020
• Renewables scenarios based on results of
  screening study (WS 2(a))
• Impact of different plant mix, more storage,
  demand-side controls to be assessed

28
Results required - 1

• Limits to renewable generation and techniques to
  overcome those limits
• System operation costs for each scenario
• Emissions of CO2, SO2 and NOx for each scenario
• Utilisation of other plant throughout the year
  (load factors, starts, typical output profiles)
• Commitment and dispatch techniques assumed for
  the model
• Dispatches for common reference points summer
  min, max, summer average, system average, winter
  average , winter peak

29
Results required - 2

• How limiting dispatches are managed
• Management of extreme events
• Impact of reducing uncertainty (e.g. by improved
  output forecasting)
• Analysis of cost differences between scenarios,
  to show key drivers of cost differences
• Plant mix, energy storage and controllable load
  impacts
• Dispatch rules emerging from studies
• Any outputs required for WS3 WS4

30
Method of Working

• Close co-operation with TSOs throughout
• Timely interim reports
• Regular consultations with stakeholders
• High quality final reports
• Presentation of findings to stakeholders

31
Timings

• Study to be let about end of March or early April
  2006
• Methodology discussions on-going
• Methodology report - mid-summer
• Test case late summer
• Draft final report late autumn

32
Key Selection Criteria

• Very well developed understanding of dispatching
  power systems
• Knowledge or experience of managing uncertainty
  and variability in power systems
• Well able to interface with TSOs, peer and
  intellectual reviewers
• Meetings with short-listed bidders may be
  required

33
Question time

• Panel

34
Wind Variability and Uncertaintythese slides are
for reference and not part of presentation
35
Variability
36
Objective function and constraints

• Objective
• Within security standards -Total dispatch cost
  minimisation (including start-up etc and fixed
  heat
• Constraints
• Ramp rates
• Time constraints and emissions limits
• Security operating constraints

37
Structure methodology

• Review dispatch methodologies in use / under
  development with high penetrations of renewables
• Agree dispatch methodology and reserve strategies
  with TSOs for
• Routine daily management
• Extreme events
• Test methodology for realism and accuracy

38
Future value factors

• Primary fuel costs
• Coal up to / tonne
• Gas up to / tonne
• gas to be unconstrained up to pipeline capacity
• Oil up to / tonne
• Carbon emissions maket price /tonne
• Nox and Sox not considered significant

39
Structure

• Scenarios deal with different renewables
  penetration and mix
• Base Cases / per scenario to be set up as a datum
  from which to measure
• Base case new plant added to the system and
  replacement plant will be CCGT
• Developed cases show quantified benefits or not
  over base cases
• Plant mix, controllable load, energy storage,
  better forecasting
• Consultant to propose optimisation of methods and
  number of studies