Doc

umen

t num

ber

Power System Modeling at the Interplay of Markets, Control and Engineering and practical Smart Grid experiencesMartijn Duvoort and Jasper FruntPisa, 28-1-2013

E-Price as smart grid enabler

Contents

Introduction- DNV KEMA- E-price vs. smart grids

How does the E-Price system function- Model description and architecture- Outputs and study example- Other model applications

How E-Price enables smart grid extensions

DNV KEMA references

2

3

Introduction

DNV KEMA Energy & Sustainability offers innovative solutions to customers across the energy value chain, ensuring reliable, efficient and sustainable energy supply, now and in the future.

2,300+ experts across all continents

KEMA and DNV combined: a heritage of nearly 150 years

Headquartered in Arnhem, the Netherlands

Offices and agents in over 30 countries around the globe

DNV KEMA Energy & Sustainability

4

The DNV Group

5

10,400employees

300offices

100countries

DNV Group

DNV KEMAEnergy &

SustainabilityDNV

Maritime Oil & GasDNV

Business Assurance

We understand:The business consequences of a technical decision,

and the technical consequences of a business decision.

Technical consultingdifferentiators

Unique client value

Management consultingdifferentiators Impartiality Deep operating

experience Proven methodologies Global utility industry

experience

Senior industry technical leaders

Insight born from experience

Operational understanding of complex component and system operations

6

Our customers span the energy value chain

‘1985 - ’20?? - De-central sys-tem, Hierarchical communication

Recources

Production Transport & Distribution Use

8

How does the E-Price system function?

Model description Includes both day ahead and real-time behavior of

- Transmission System Operator, and- Balance Responsible Parties

DA market operation based on APX DA spot market algorithm

Model of the Netherlands in European power system and includes cross-border interconnections with their limitations

Physical model is based on active power balance and includes dynamics of power plants (ramp rates, delays, etc.) and grid

9

Predictions Trading at spot market Bidding AS Unit

commitmentEconomicdispatch

Dynamicstudy Settlement

Integration of day ahead and real-time models

10

Simulation frameworkDay ahead (DA) Real-time (RT)

TSO DA

Energy market

AS market

BRP DABRP DABRP DA

BRP Unit Commitment

BRP Energy market bidding

BRP AS market bidding

BRP DABRP DABRP RT

BRP DABRP DAProsumers

RT

TSO RT

Physical power system

Calculate Performance

metrics

Case Studyresults

Input data sets

Cross-border schedules,

E-Programs,AS

Allocations

Integration of day ahead and real-time sections

Integration of markets and physical power system

Physical model Based on DNV KEMA’s KERMIT model which is fine tuned and verified on real-life

(Dutch Power System) measurements

11

Standard Inputs:LoadPlant SchedulesGeneration PortfolioGrid ParametersMarket/Balancing

Outputs:ACEPower Plant MW OutputsArea InterchangeFrequency Deviation

Scenarios:Increasing Wind Adding ReservesStorage ParametersTest AGC ParametersTrip Events

KERMIT 24h Simulation

Generation• Conventional• Renewable

Inter-connection

Frequency Response

Real-Time Market

Generator Trip

Wind Power Forecast vs. Actual

Load Rejection

Volatility in Renewable Resources

Physical model and power balance

Imbalance caused by- Trip of generation or load, or separation of areas.- Control strategy- Prediction errors in load or generation

Balance should always be restored

12

50 Hz<50 Hz

frequ

ency

Primary

Secondary

Inertia

0 30 900

t [s]

t [s]

pow

er

0 30 900

Deployment of balancing capacity

Frequency response

System response during unit trip

44.5

45

45.5

46

46.5

Plan

ts o

n []

# plants on

-1000

-500

0

500Po

wer

[MW

]

Imperfect forecast error

0 1 2 3 4 5 6 7 8x 104

-100

0

100

Time [s]

Pow

er [M

W]

Wind fluctuations from perfect forecast

Imbalance definitions Imbalances have been defined to validate the algorithms

- Power plant trips

- Wind power forecast errors

- Wind power fluctuations

13

Model outputs

Delta frequency

Power realized vs. scheduled on

cross-border interconnection

BRPs reference power vs.

E-program vs. AGC set points

BRPs reference power vs. realized

ACE and PACE, action TSO

Load

Etc…

14

Observations – hourly frequency excursions

Generator dispatch

Observations of the current model:- Frequency disturbances at PTU-crossings

Ramping behavior of generators causes frequency excursions

15

0 :00 5 :00 10:00 15:00 20:0049.9

49.95

50

50.05

50.1

Freq

uenc

y [H

z]

Time [hh:mm]

0 5 10 15 20

12

14

16

18

20

Time [h/day]

Pow

er [G

W]

Schedules

Production ScheduleLoad

Observations - trip studies Power plant trip of approximately 130 MW

Primary response not affected by double sided market

Double sided market gives a faster recovery of frequency (green/yellow)

Double sided market with prosumers gives a faster recovery (orange/red)

16

5.75 5.76 5.77 5.78 5.79 5.8 5.81x 104

-0.015

-0.01

-0.005

0

Freq

uenc

y de

v [H

z]

Time [s]

Trip at 57600 seconds

CS1CS16CS21CS26CS31

5.75 5.76 5.77 5.78 5.79 5.8 5.81x 104

8100

8200

8300

8400

8500

Gen

erat

ion

NL

[MW

]

Time [s]

Trip at 57600 seconds

CS1CS16CS21CS26CS31

5.75 5.76 5.77 5.78 5.79 5.8 5.81x 104

5000

5050

5100

5150

5200

Impo

rt [M

W]

Time [s]

Trip at 57600 seconds

CS1CS16CS21CS26CS315.76 5.77 5.78

Time [s]

Trip at 57600 se

Observations - double sided market

Double sided market reduces control effort

Causing imbalance is penalized more in double sided markets

BRPs are more conservative, more reservations for AS

17

Observations - prosumers

18

Prosumers reduce imbalance energy

Prosumers have no effect on primaryresponse

Prosumers lead to a faster frequencyrecovery after a trip

0 1 2 3 4 5 6 7 8x 104

0

200

400

600

800

1000

1200

1400

1600

Time [s]

Cum

ulat

ive

Win

d Po

wer

[MW

]

Wind farm 1Wind farm n

Other model applications

Trip studies

Testing of AGC algorithms, double sided market

Testing of (real-time) dispatch algorithms- Improved dispatch with RES in portfolio- Over multiple program time units with

improved flexibility

Observing market behavior of market parties

Determine the effects of energy storage and demand response

19

20

How does this enable smart grid extensions?

Tomorrow: Smart Grids

Reforming

Gas

Electricity

Multi-directional power and information flows

Definition:“A smart system, where generation and demand are balanced on a local level to prevent grid congestion.”

Balancing signal is a local priceincentive

What is a local price?

22

Plocal = Pgrid incentive + Psystem

Pwholesale+ Pbalance

DA – intraday real-time

E-Price real-time price

Purpose of Smart grid Management of energy use

- Shifting load based on availability of generation and grid and price- Automated with smart appliances, storage and through demand/load management software

Allowing of micro trading- Sell own generation (PV, µCHP, small wind) and storage (EV, H2, batteries) to anyone

anywhere- Hire distribution capacity

Lower operational cost and capex- Remote diagnosis, condition based maintenance, embedded systems, load management

More market based balancing, less control effort

23

Challenges for smart grids Solved by E-Price::

- Good relation between local power price and system price- Not only day-ahead wholesale price, but real-time price including balancing component

Open problem:- Solve “hierarchy” problem – how to influence local power flows in a mixed grid topology?

24

Hierarchical communicationLocation parameterMeshed ownership

25

DNV KEMA references

Power Matching City: Smart Meter Allocation

Wholesale processes need to be modified in order to facilitate smart energy concepts

such as virtual power plants, real time pricing and time of use contracts for small

consumers / prosumers.

26

In cooperation with a Dutch energy supplier and distribution company

• Modified processes are designed, implemented and validated in a test environment

• Results are shared with the Dutch energy sector and government.

Day + 1 / Day + 4 / Day + 9

Allocation Hoogkerk

“Sm

art M

arke

t”E

ssen

t BR

PE

ssen

t MD

CE

nexi

sH

ome

P3

See: SAU

Deter-mination

Smart meter reading

Allocation

Aggregation validation

DAACPE-66

Imbalancemanagement

DAACPMSCONS

Allocationcheck

E150

Meter data validation

Interval meter reading

Meter data estimation and

aggregation

P4

Interval meter read CTS

Notify missing data

= Production Environment = Modified process / interface

Interval meter reading Repo

sitory

P1Home Agent

Interval meter reading

Sanity check

MCFMSCONS

Smart Energy Collective The development and testing of a set of integrated, interoperable smart energy services with

corresponding technologies and infrastructures.

The set up of an open innovation platform with accompanying organization where these integrated, smart energy products and services can be developed, demonstrated, and tested on a large scale on different locations. Approx. 10 locations will be developed with a total of approx 5000 smart grid connections.

The development of a common market for smart energy services with sufficient volume.

Participating: DSOs, TSOs, Energy suppliers, ICT system integrators (Humiq, Logica, IBM, Sogeti), Project development (Heijmans), Financial sector (Rabobank), Installation company (Imtech, Unica), Technology provider (Philips, Nedap, ABB, Siemens, Heliox, Bosch), Product industry (Daalderop, Itron, Miele).

DNV KEMA Smart Grid Cost Benefit Model

Integrated approach to evaluate (avoided) costs for transmission grids,

distribution grids, central electricity generation and imbalance services.

28

Energy storage

Conventional power stations

Electric vehicles

Solar PV

Wind turbines

Commercial buildings

Industry

Dwellings

virtual power plantDispersed

generation (CHP)

Smart meter

Pow

er

Time

Solar PV

Wind

• Based on actual load profiles

• Includes advanced technologies (micro-CHP, electric vehicles, heat pumps, etc.)

• Includes effects of electricity storage

• Multiple scenarios possible

Demonstrator: EU ADDRESS project

Objectives:

Enabling active demand

Exploitation of benefits from active demand

Active customer

participation

Increased power system efficiency

Integration of Renewable Energy Sources

30

GROW-DERS

Demonstration of grid connected storage systems (Li-ion batteries, flywheels)

Development of an assessment tool to determine optimal storage applications

Determine where grid connected storage is most attractive

Focus on distribution grid (LV)

EU under 6th framework programme

31

Large-scale energy storage

Our contribution• Innovator

• Initiate project and set up partnership structure

• Feasibility study

• Artificial island in North Sea (The Netherlands)

• Store (wind powered) electricity when demand is low – e.g. at night – and make use of it during the day.

www.dnvkema.com

www.dnv.com

© KEMA Nederland B.V., Arnhem, the Netherlands. All rights reserved.It is prohibited to change any and all versions of this document in any manner whatsoever, including but not limited to dividing it into parts. In case of aconflict between the electronic version (e.g. PDF file) and the original paper version provided by KEMA, the latter will prevail.KEMA Nederland B.V. and/or its associated companies disclaim liability for any direct, indirect, consequential or incidental damages that may result fromthe use of the information or data, or from the inability to use the information or data contained in this document.The contents of this report may only be transmitted to third parties in its entirety and provided with the copyright notice, prohibition to change, electronicversions’ validity notice and disclaimer.