distributed and hierarchical congestion management in ... · for monitoring & control systems...

ideal grid for all

Distributed and hierarchical congestion management in distribution networks containing distributed energy resources

Tutorial in ISGT Europe 2016 conference

Ljubljana, Slovenia

Sunday October 9th 13:30-15:00

Professor Sami Repo, Tampere University of Technology

Senior Scientist Anna Kulmala, VTT Technical Research Centre of Finland

IDE4L is a project co-funded by the European

Commission (Project no: 608860)

ideal grid for all 10/04/2017SLIDE 2

• Active Network Management concept

• Distributed automation system

• Congestion management concept

• Demonstrations

• Conclusions

WWW.IDE4L.EU – ISGT Europe 2016 tutorial October 9th 2016

ideal grid for all

The Project at a glance

• FP7 demonstration project (9/2013 – 10/2016)

• Scope: Active distribution network management• From the planning to the real-time operation

• From the MV up to LV single customer

• DSO interaction with TSO, DER, μGrid and Aggregator

SLIDE 3 10/04/2017 WWW.IDE4L.EU – ISGT Europe 2016 tutorial October 9th 2016

ideal grid for all

Breakthroughs

18/05/2016SLIDE 4

WP2Planning tools for distribution

network management

ANM concept

Target and expansion

planning including ANM

Operational planning including DER uncertainty

WP3 Distribution

network automation architecture

Automation concept

Smart meter as a sensor

Testing Platform for monitoring & control systems

Hierarchical and decentralized automation

WP4 Fault location, isolation and

supply restoration

Decentralized FLISR

IEC 61850 Distribution

Protection System Reconfiguration

Microgridinterconnection

switch

WP5Congestion

management

Decentralized state estimation and state forecast

Tertiary control –Network

reconfiguration

Secondary control – Coordination of voltage controllers

Dynamic tariff

WP6Distribution

networks dynamics

Aggregator concept

Optimal scheduling of flexibility

Transmitting synchro-phasors &

real-time model syntheses

Improved microgridoperation

WP7 Demonstrations


Concept Monitoring Control Market

ideal grid for all

MicrogridsEnergy communities

Vision of future smart grid

System management and design

Advanced monitoring

Market design

Aggregator

Smart charging of EV Smart homes and PV

Storage

Controllable loads and energy efficiency

Power to gas

BalancingDistribution automation

Renewable energy resources

Grid infrastructure

ideal grid for all

Policies of electricity network

• Today networks are always over-dimensioned due to quality of supply obligations and missing possibility to control DERs

• Some companies are already forced to utilize production curtailment to manage their networks

• In future more flexibility is needed to integrate more RES and DERs in power system

• Controllability of distribution network via advanced ICT

• Decentralization of network management due to scale of the system

SLIDE 6 WWW.IDE4L.EU – ISGT Europe 2016 tutorial October 9th 201618/05/2016

ideal grid for all

Active distribution network• Active distribution network utilize DERs in grid management Active Network Management (ANM)

• DERs are integrated as part of grid and markets instead of ”fit and forget” connection

• ANM should provide synergy benefits for DSO and customers

18/05/2016 WWW.IDE4L.EU – ISGT Europe 2016 tutorial October 9th 2016SLIDE 7

ideal grid for all

IDE4L automation architecture

18/05/2016SLIDE 8 WWW.IDE4L.EU – ISGT Europe 2016 tutorial October 9th 2016

Commercial

aggregator

systems

DSO

Energy retailer

Market

operator

platform

Balance

responsible party

TSOAMR HUB

Control HUB

Weather HUB

DG

IED

DR

Weather

data

Enterprise Service Bus

DMSMDMS CISNIS

SCADA NIS

Substation automation

Secondary substation

automation

IED

IEDRTU

Secondary substation

automation

IED

PMU

PMU PQ

Smart

meters

Smart

meters

DER automation RTUDER automation

Commercial aggregator

IED

DG

IED

DR

IED

ideal grid for all

Roles of grid operatorsand aggregator

1. DSO/TSO• Validates the submitted offers:

• Off-line validation

• Real-Time validation

• Purchases flexibility services for avoiding network constraints

• Calculates and provides the Flexibility Table (Limits for each Load Area)

2. Aggregator• Forecasting of consumption, production, price, etc.

• Flexibility estimation of customers

Determination of market bids

• Commercial optimal planning

Maximization of aggregator profit


ideal grid for all

Monitoring, protection and control system• Complete network will be monitored and controlled

• Intelligent Electronic Devices (IEDs)

• Coordination and merging of information and decisions at substations

• DA applies variety of communication technologies• Primary substations - SCADA and possibly other IT systems (fibre optics,

wireless)

• Secondary substations and MV switching stations (wireless)

• Smart meters (PLC or wireless)

• Ethernet is becoming the prevalent communication standard for all automation devices

• IEC 61850 GOOSE and MMS

• DLMS/COSEM

• IEC 60870-5-104

• Modbus/TCP over LAN/WAN


ideal grid for all

Control of DERs from DSO’s viewpoint• Regulation

• Connection requirements technical capabilities for the control of DERs

• Dynamic tariffs to incentivize load shifting• Retail off-peak day-ahead prices

• Grid off-peak network load

• Direct control• DSO’s own resources (OLTC, Reactive power compensation and

FACTS)

• Contracted non-market based control, e.g. voltage control of DG units

• Emergency control to act just before protection

• Flexibility services from Commercial Aggregator• Scheduled re-profiling of flexible DERs

• Conditional re-profiling of flexible DERs


ideal grid for all

Active network planning

• Active network becomes alternative for network reinforcement

• Postponing investments of physical infrastructure by ANM

• Replacing network reinforcement with smart functionalities

• Traditionally worst case design principle• Firm connection capacity always available for all customers

• DG impact maximum production – minimum loading condition

• Leads to over-dimensioning of network and the evaluation of smart functionalities is limited to peak conditions

• Stochastic planning of active network• Non-firm connection (based on dedicated contract) increase network

hosting capacity remarkably

• Enable full utilization of ANM


ideal grid for all 10/04/2017 WWW.IDE4L.EU – ISGT Europe tutorial October 9th 2016SLIDE 13




• Demonstrations

• Conclusions

ideal grid for all

General architecture

10/04/2017 WWW.IDE4L.EU – ISGT Europe 2016 tutorial October 9th 2016vSLIDE 14

ideal grid for all

Motivation

SLIDE 15

• Background• Little monitoring penetration in the

distribution grid:• primary substations

• secondary substation remote control

• average values only

• Electronic meters deployment mainly for billing and CRM

• Needs to be developed• LV grid: EV, PV, HP and demand-response

schemes mainly affect the LV grid

• Secondary susbtations: Improve the monitoring, protection and control of secondary substations

• TSO-DSO: Need of dynamic information from DSO grid

10/04/2017 WWW.IDE4L.EU – ISGT Europe 2016 tutorial October 9th 2016

ideal grid for all

Pillars

SLIDE 16

• Amount of data increases • 10 of PSs -> 106 customers

• averages only -> high-frequency measurements

• Decentralized approach• Data is collected/processed locally (LV

data -> in SS; MV data -> in PS)

• Only summary/alarms are reported to upper levels

• Benefit: impact on CAPEX (scalability / modularity)

• Less-demanding communication is required

• Faster response

• reuse of existing automation components


ideal grid for all

Pillars

SLIDE 17

• The number/type of monitoring devices are increasing

• Smart meters

• Fault detectors / protections

• Power quality meters / PMU

• Standards are needed to:• Limit the integration time

• Limit the maintenance time

• Main standards:• CIM for grid assets

• 61850 for data about the grid

• DLMS/COSEM for metering data

• Benefit: Positive impact on CAPEX and OPEX (interoperability)


ideal grid for all

Commercial aggregator market setup


Purchase/Sale bids

Commercial Aggregators

Wholesale Market

Other market

participants

Bilateral contracts

Flexibility market

Sale bids

TSO

Flexibility activation requests

Flexibility procurement

DSOs

Flexibility procurement


Flexibility validation

Consumers Prosumers


TSO

/DSO

co

ord

inat

ion

Market clearing

Market clearing

Market clearing

Market clearing

PurchaseSale bids

ideal grid for all

Market operator (MO) as neutral market facilitator



Flexibility productsTwo types of standardized Flexibility Products

AD Product Conditionality Example

Scheduled re-

profiling (SRP)

Unconditional

(obligation)

The aggregator has the obligation to

provide flexibility services

Conditional re-

profiling (CRP)

Conditional

(real option)

The aggregator must have the capacity

to provide flexibility services


ideal grid for all

Aggregator’s real time algorithm


DSO

Aggregator

DERs

1

2 3 7

5

2 64 8

• 1: Order of flexibility activation

• 2: Status request (baseline + available flexibility) and operational status return

• 3: Allocation of flexibility volume among DERs proportional to their available amounts

• 4: DERs’ activation

• 5: Activation confirmation

• 6: Energy measurements (request and return)

• 7: Modification of flexibility volume target for remaining activation period and re-allocation of set points among DERs

• 8: Sending of modified control set points

ideal grid for all

Comparison of architectures

Distributed / Decentralized

• Local challenges

• Small-scale resources

• Scales well to whole system

• Less dependent on communication

• Needs to be coordinated with centralized systems

• Supports novel ideas: active networks and microgrids

• Information exchange is the key question for successful implementations

Centralized

• Global challenges

• Large-scale resources

• Scalability is an issue

• Very dependent on communication

• Does not necessarily require existence of distributed ach.

• Less flexible for novelties

• Software integration and communication QoS are the key questions


ideal grid for all

Substation Automation Unit (SAU)


ideal grid for all

Use cases

SLIDE 24

IDE4L hierarchical network management

… a (very) simplified description …

1. Monitoring• Real-time monitoring

• Load and production forecasting

• State estimation and forecasting

• Dynamics of distribution grid

2. Protection• Logic selectivity

• FLISR with DERs and μGrid

3. Control• Congestion management

• Optimal scheduling

• μGrid voltage control

• Dynamic grid tariff

Monitoring

CB, OLTC

Short-term forecast

State estimation

Secondary control

IED, Primary control

CB, DER

IED, Primary control

Tertiary control

Long-term forecast

Day-ahead market

Real-time/intra-hour

Day-ahead and Intra-day

Hard real-time

e.g. fault location

e.g. V/VAr regulation

e.g network reconfiguration

Secondary control

Flexibility market

Commercial aggregator

e.g. DER scheduling

e.g. CRP and SRP signals

Market

Tertiary controller

Secondary controller

Monitoring Estimation

Forecasting

Primary controller

Primary device


ideal grid for all

Substation Automation Unit


SAU

Interfaces

Database

Applications

DLMS/COSEM

MMS

Prosumers – Smart meters and DER IEDs

Control Center - DMS

Substations - IEDs

IEC 61850

CIMManagement

model

Bridge model

State Estimation

Power control

ForecastData Acquisition

Reports

ideal grid for all

DemonstrationsSLIDE 26

Testing phase: a three-step procedure

Building-blocks, e.g.:

1. Algorithms2. Protection devices3. Third party devices4. Third party software

Groups of building-blocks, e.g.:

1. State estimation algorithm within a PC connected to an RTU via a 61850 interface

Use cases, e.g.:

1. Monitoring of LV grid (PC + state estimation + RTU + Smart meters + interfaces)

1stDev. lab 2nd

Integration. lab 3rdDemo


ideal grid for all

Unareti Demo Site Overall Architecture


BRESCIACONTROL CENTER

PRIMARY SUBSTATION AUTOMATION UNIT

PROTECION SYSTEMS

PRIMARY SUBSATION SCADA SYSTEMS

SMART METER

PRIMARY SUBSTATION

LEVEL

SECONDARY SUBSTATION

LEVEL

LOW VOLTAGE NETWORK LEVEL

PHOTOVOLTAIC PANEL

STORAGE

MV Demonstrator

LV Demonstrator

SECONDARY SUBSTATION AUTOMATION UNIT

PROTECTION SYSTEMS

BREAKERS

ideal grid for all

LV Network Demonstrator


ideal grid for all

Net load forecast of prosumers


Ostkraft Fenosa, producer

Unareti

ideal grid for all

State estimation


Ostkraft

Unareti

ideal grid for all

Design and implementation of SAU


ideal grid for all 18/05/2016SLIDE 32 WWW.IDE4L.EU – ISGT Europe 2016 tutorial October 9th 2016

1. Use Cases

2. SGAM Architecture

3. Implementation of architecture

4. Architectureevaluation


1. Use Cases

MonitoringUse Cases

Control Use Cases

BusinessUse Cases

• State estimation, forecast, network update, measurement collection

• LV, MV, control center power control, block OLTCs, FLISR

• Purchase of energy and flexibility, activation of flexibility



1. Use CasesUse Case description - example

Steps Information producer Information receiver Function Information exchanged Requirement

1

2 … … … …

Information exchanged

Switch statusVoltage measuremenCurrent measurementPower/energy

measurement

Functions

Data acquisition

Data report

Data storage

Signals sampling

Statistical calculation

ActorsSAU(PSAU).MMSSAU(PSAU).RDBMSSAU(PSAU).IEC104DMS.MMSSensorSAU(SSAU).MMSDMS.ModbusSAU(PSAU).FunctionsSAU(PSAU).ModbusDMS.IEC104IED(PSIED).MMS

IED(PSIED).functions

Requirements

Requirement: transfer time

Requirement: Transfer rate

Requirement:

Synchronization

Requirement: Availability

SAU(PSAU).MMS SAU(SSAU).MMS Data Report Switch Status TT = 100 ms …


ideal grid for all

2. SGAM architectureGeneral description and link to use cases

18/05/2016SLIDE 35

1. Business framework

2. Functions to be implemented

3. Data models in the main automation standards

4. Communication protocols

5. Components, both hardware and software to take part to the automation system

Smart Grid Coordination Group, CEN-

CENELEC-ETSI, Tech. Rep., 2012


ideal grid for all

2. SGAM architectureBusiness layer and mapping to component layer

18/05/2016SLIDE 36

Business layer

• Business actors are connected by business transaction

• Each one has a business goal

• Business actors are mapped onto components

Business goals

Transmission System Operator

(TSO)

Service Provider (SP)

Market Operator (MO)

Commercial aggregator (CA)

Distribution System Operator (DSO)

Retailer

Prosumer

Power schedule and activation price

Business transactions


DSO

Prosumer

Transmission System Operator (TSO)

Service Provider (SP)

Market Operator (MO)

Commercial aggregator (CA)

TSO Energy management system

Service provider platform

Market Operator platform

Commercial aggregator system (CAS)

Intelligent Electronic Device (IED)


MicroGrid Central Controller (MGCC)

Distribution Management System (DMS)

„Business Actors“ „Automation Actors“

Ensure market settlement

Maximize income from energy selling

and purchase

Monitor of grid status, power quality and

sequrity requirements

Sell services as price, weather, generation

forecasts

Optimize energy costs of portfolio

ideal grid for all

2. SGAM architectureComponent layer

18/05/2016SLIDE 37

Sensor Actuator

Transmission System Operator Energy Management System

(TSOEMS)

Service Provider Platform (SPP)

Market Operator Platform (MOP)

Commercial aggregator automation system (CAAS)





0..*

1..*1..*1..*

1..*

0..* 0..*

1..*

1..*

1..*

0..*

0..*

1..*

1..*0..*

Substation automation unit (SAU)

MMS

Relational database managment system (RDBMS)

Time series database (TSDB)

Interfaces

Database

Functions

SAU

Modbus

· Data report· Data storage· Check flag· CIM parsing· Data reconstruction· Detect error· Fault isolation· Load forecast· Missing data· Optimal power flow· power quality control· power quality indexes· algorithm performance index· Second fault isolation· State estimation· State forecast· Statistical calculation· Data acquisition· Protection update· Reading/Writing IEDs setting

DLMS/COSEM

IEC 61850-90-5

Web Services (WS)

Each Automation actor is defined in terms of

• Interfaces

• Database

• Functions

New automation actors developed :

• Substation Automation Unit

Further development for

• Commercial aggregator

• Distribution management system (DMS)

• MicroGrid Central Controller

Also present in function

layerand UCs


ideal grid for all

Bid acceptance/modification

Bid submission

Check flag

CIM parsing

Commercial optimal planning

CRP activation request

CRP Validation request

Data acquisition

Data Curation and Fusion

Data reconstruction

Data report

Data storage

Detect error

Dynamic info derivation

Fault detection

Fault isolation

Load area configuration

Load forecast

Market clearance

Market infos

Missing data

Non-convergence detection

open/close switch

Optimal power flow

power flow

power quality control

power quality indexes

Protection update

algorithm performance index

Reading/Writing IEDs setting

Second fault isolation

Signals sampling

State estimation

State forecast

Statistical calculation

synchronization

Validation reply

Validation request

2. SGAM architectureFunction layer

18/05/2016SLIDE 38

Functions are mapped to

• Actors

• Use cases

• Zones and domains of Smart Grid Plane

LV real time Monitoring


SAU

Function list

MV real time Monitoring

MV State Estimation

Control Center Power Control

MV Power Control

DMS

MGCC

IED

ideal grid for all

2. SGAM architectureInformation layer

18/05/2016SLIDE 39

class Business Context View -PSAU-SSAU

Components::PSAU

Computer

Components::SSAU

Computer«Information Object» Network static data

«Information Object Flow»

«Information Object» OLTC control


«Information Object» State Estimate/forecast


«Information Object» State Estimate/forecast


«Information Object» Current/Voltage Meas., «Information Object» Meas. statistics,

«Information Object» Power/Energy Meas., «Information Object» switch status


IEC 61850 – data models Logical node Data object Data Attribute

ATCC BndCtr ASG setMag AnalogueValueCIM – data models


class Business Context View -SSAU-IED

Components::D

:IED

Components::SSAU

Computer

Components::

SS :IED

Components::MGCC

Computer

Components:

:D :PMU

Components:

:DIED :

Interconnection

Switch

Components:

:DIED :

Smart

meter

Components:

:DIED :

SW-CTRL

Components:

:SSIED:

MDC

Components:

:SSIED :

SW-CTRL

Components:

:SSIED :

AVC-CTRL

Components:

:SSIED :

PMU

Components:

:PS :PMU

Components:

:SSIED :

RTU

«Information Object» Setting values of IEDs


«Information Object» Setting values of IEDs


«Information Object» Current/Voltage Meas.,

«Information Object» Frequency Meas.



«Information Object» Frequency Meas.



«Information Object» Power/Energy Meas.,

«Information Object» switch status










«Information Object» flags


«Information Object» status polling


class Canonical Data Model

Generation Transformation Distribution DER Customer Premise

Market

Enterprise

Operation

Station

Field

Process

Name: Canonical Data Model

Author: chn

Version: 0.3

Created: 17.6.2013 14:14:17

Updated: 1.7.2015 9:50:42

Canonical

Data

Model

«Data Model Standard»

CIM/XML


IEC 61850-7-4/ IEC 61850-7-420


DLMS/COSEM

ideal grid for all

2. SGAM architectureCommunication layer

18/05/2016SLIDE 40

Transmission System Operator Energy Management System

(TSOEMS)

Service Provider Platform (SPP)

Market Operator Platform (MOP)

Commercial aggregator automation system (CAAS)





Tech 11 Tech 11

Tech 11

Tech 11

Tech 11

Tech 11 Tech 1

Tech 10

Tech 11Tech 1

Tech 1Tech 1Tech 1

Tech 1

Tech 1

Tech 11

Steps Information

producer

Information

receiver

Function Information

exchangedRequirement

1 SAU(PSAU).MMS

SAU(SSAU).MMS

Data Report

Switch Status Transfer Time = 500 msTransfer Rate = 1000 kb/sSynchronization accuracy = … Availability = …

2 … … … …



Function layer

• Functions have been realized in WP4 (FLISR), WP5 (Monitoring and LV, MV, control center control), WP6 (business and commercial aggregator)

• Functions are adapted in order to read and write from a standardized IDE4L database

Smart Grid Coordination Group, CEN-

CENELEC-ETSI, Tech. Rep., 2012

Communication layer

• The requirements of the information exchange grouped onto technology classes

2. SGAM architectureOne step toward implementation

Information layer

Exchanged data are clustered onto classes and mapped to

• CIM data models for static data and business related data

• 61850 for real time data

Component layer

Each component is implemented as

• Software/Hardware interfaces

• Database

• Functions


ideal grid for all

3. Implementation of architectureDatabase structure

18/05/2016SLIDE 42

Measure & Command Model

• Physical devices, logical devices, logical nodes, data objects and data attributes with real time data

• Set of information to parameterize the communications interface to each physical device (such as IP addresses, TCP ports, users and passwords, etc.)

Measure&CommandModel

NetworkModel

BridgeModel

ManagementModel

Management Model

• Represents the models related to an algorithm.

• instantiate, parameterize and control the execution of a specific algorithm

Bridge ModelIt is the connection schema for all other schemas. • Relations among Measure & Control

real time quantities with the network topology

Network Model• Network topology and parameters of lines,

customers and generators


ideal grid for all

Conclusions and Exploitation of IDE4L architecture

18/05/2016SLIDE 43

1. Use cases

• 29 use case detailed descriptions (11 monitoring, 11 control, 7 business)

• List with description of actors, information exchange, functions and communication requirements

2. SGAM architecture

• SGAM communication, information, component, business detailed SGAM layers in .xls and enterprise architect files

3. Architecture Implementation

• 61850, CIM information mapping for whole set of information exchanges in UCs in .xls tables to facilitate standard implementation of architecture

• Database structure and sample communication interfaces

4. Evaluation of architectures

• Architecture metrics definition and results






• Demonstrations

• Conclusions


HV/MV MV/LVMV feeder LV feeder

Voltage

1 pu

Permissible

voltage

variation

Allowable

range of

substation

voltage

Minimum load, substation voltage at its highest value

Maximum load, substation voltage at its lowest value

Transformer

tapping boost

Transformer

voltage drop

LV feeder

voltage drop

MV feeder

voltage drop

Voltage drop margin

Voltage rise

margin

Voltage profiles without generation

Why congestion management?




Voltage

1 pu

Permissible

voltage

variation

Allowable

range of

substation

voltage



Transformer

tapping boost

Transformer

voltage drop

LV feeder

voltage drop

MV feeder

voltage drop

Voltage drop margin

Voltage rise

margin


Voltage

1 pu

Permissible

voltage

variation

AVC relay

dead band

(2 DB in

Figure 3.2)

Minimum load, maximum generation, substation voltage at its highest value

Maximum load, no generation, substation voltage at its lowest value

Maximum load, maximum generation, substation voltage at its lowest value

Voltage outside the permissible range

Voltage profiles

with generation

Allowable

range of

substation

voltage




Voltage

1 pu

Permissible

voltage

variation

Allowable

range of

substation

voltage



Transformer

tapping boost

Transformer

voltage drop

LV feeder

voltage drop

MV feeder

voltage drop

Voltage drop margin

Voltage rise

margin


Voltage

1 pu

Permissible

voltage

variation

AVC relay

dead band

(2 DB in

Figure 3.2)

Minimum load, maximum generation, substation voltage at its highest value

Maximum load, no generation, substation voltage at its lowest value

Maximum load, maximum generation, substation voltage at its lowest value

Voltage outside the permissible range

Voltage profiles

with generation

Allowable

range of

substation

voltage

Active congestion

management methods

decrease the total costs

of the network in many

cases


Control hierarchy of congestion management

• Secondary control manages resources in one MV or LV network

• Mitigates congestions and optimizes the network state

• Operates through changing the set points of primary controllers

• Based on state estimation results

• Tertiary control utilizes forecasts for a longer time interval and manages the whole distribution network

• Network reconfiguration

• Real power control through the market place

• Dynamic tariff

Real and reactive

power controllers

AVC relay

SSAU

SSAU

MV/LV

HV/MV

Load

controller

ACDC PV

Converter

controller

MV/LV

... LV

Compensator

controller

Tertiary control

Set points

Database

Load and

production

forecast

State

estimation

Secondary

power control

Interfaces to primary controllers, measurement

devices, SSAUs and control centre

...Algorithms

Smart meter

Measurement data

PSAU

Control centre

Switch

controller

DMS/SCADA Market place


Interactions of the congestion management system


• Load and production forecast, state estimation and real time power control algorithms are functions of the SAU and are responsible for monitoring and control of either one MV or one LV network

• All information exchange between the functions is realized through the database

• Monitoring and load and production forecast functions are executed asynchronously

• State estimation and power control execution is synchronized by using database flags

Real time monitoring and secondary control


• The load and production forecast algorithms are time series forecasteralgorithms

• The state estimation algorithm is a weighted least squares state estimator that uses branch currents as state variables

• The power control algorithm is an optimal power flow (OPF) algorithm implemented using sequential quadratic programming (SQP) algorithm

• The objective function is formulated to minimize network losses, production curtailment, load control actions, the number of tap changer operations and the voltage variation at each node.

Real time monitoring and secondary control


Network reconfiguration and market agent algorithms

• The network reconfiguration and the market agent algorithms are executed sequentially

• Both are optimizing algorithms: network reconfiguration utilizes genetic algorithm and the market agent primal/dual interior point algorithm





• Demonstrations

• Conclusions


Lab Demo Site

Field Demo Site


Lab Demo Site

Field Demo Site

Secondary controldemonstrations


Demonstrated in laboratories and real distribution networks


Lab Demo Site

Field Demo Site

Verification of correct algorithm operationsimulation sequences planned to test the algorithm in extreme conditions• MV and LV secondary

control


Lab Demo Site

Field Demo Site

More realistic simulation sequences• MV and LV secondary

control


Lab Demo Site

Field Demo Site

PV reactive power controllable• LV secondary control

PV real power controllable• LV secondary control


Lab or Simulation Demo Site

Tertiary controltesting


Lab or Simulation Demo Site

Dynamic tariff

Market agent and aggregator functionalities

Network reconfiguration and market agent

ideal grid for all

Secondary control

10/04/2017 WWW.IDE4L.EU – ISGT Europe tutorial October 9th 2016SLIDE 62


RTDS simulations atTUT laboratory

• RTDS emulates the distribution network

• Real IEDs and a real SAU connected to the simulation


RTDS simulations atTUT laboratory

• RTDS emulates the distribution network

• Real IEDs and a real SAU connected to the simulation

• The simulation network is a reduced model of the real Unareti LV network

• 6 controllable PV generators (size increased from the real ones)

• Tap changer at the MV/LV-transformer (not available in the real network)


Time DG output Loading

0 s 100 % MIN

50 s 30 % MIN

230 s 30 % MAX

410 s 100 % MAX

Generation

decreases

Load

increasesGeneration

increases

𝑓 = 𝐶𝑙𝑜𝑠𝑠𝑒𝑠𝑃𝑙𝑜𝑠𝑠𝑒𝑠 + 𝐶𝑐𝑢𝑟Σ𝑃𝑐𝑢𝑟

Operation of the LV secondary control



Generation

decreases

Load

increases

Generation

increases

Objective function

𝑓 = 𝐶𝑙𝑜𝑠𝑠𝑒𝑠𝑃𝑙𝑜𝑠𝑠𝑒𝑠 + 𝐶𝑐𝑢𝑟Σ𝑃𝑐𝑢𝑟

+ 𝐶𝑡𝑎𝑝𝑛𝑡𝑎𝑝 + 𝐶𝑉𝑑𝑖𝑓𝑓Σ 𝑉𝑖,𝑟𝑒𝑓 − 𝑉𝑖2

ideal grid for all

Testing interactions between MV and LV secondary control


1341

1006

603

1512

117

585

60

1070

PL1

PL2

2093

2137

1136

PS0023

PV

2

PV

PV

SS1056

1056

PVPV

PV

PV

MV

LV

1

3

4

5

6

7

8

9

10 11

12

13

14

MV/MV

MV/LV

ideal grid for all

Adverse interactions do not usually occur graded time operation of AVC relays adequate to prevent hunting behaviour


Acceptable

voltage range

PV production

0.1 1.0 pu

PV production

1.0 0.5 pu

Voltages restored to

an acceptable level

Voltages too high

Changes in one LV network do not cause significant changes to MV network state MV

network secondary control does not change the set points of MV network primary controllers

Voltage levels above nominal to minimize losses

Reactive powers set to

minimize reactive power transfers

MV/LV tap changer operates

The full capacity of PV unit 14 inverter

used for real power generation

ideal grid for all

Testing interactions between MV and LV secondary control


Time Feeding network voltage

0:00 23.0 kV

1:45 23.5 kV

3:45 23.0 kV

5:15 23.3 kV


RTDS simulations at RWTH laboratory

0545

1464

1073

1341

1006

0468

0827

1354

1340

0603

0952

1056

0732

1462

0297

1190

1070 0749

1180

0585 0937

0870 1143

0604

0060

1187

0987

1136

FD8SC1

OLTC

MV / LV

FD8SC2 FD8SC3

FD7SC1 FD7SC2 FD7SC3

FD3SC1 FD3SC2 FD3SC3

FD2SC1

FD2SCe

FD2SCw

FD1SC1 FD1SCe

FD1SCw

OLTC

HV / MV

1217

1023

1024

0338

0145

1512

0378

0605

1438

0117

Rack 1

Rack 3Rack 2

Rack 4

Feed

er 1

Feed

er 2

Feed

er 3

MV and LV Unareti network modeled in the test scenarios

Rack 2

Rack 1

Rack 3

Rack 4

MV and LV grid model implemented for RTDS simulation

ideal grid for all

RTDS simulations at RWTH laboratory


Laboratory Computer IED

Oracle VM (Ubuntu 14.04) VIEDs

SM

PMUs

Analog connection

Grid

Digital connection

Analog connection

Laboratory Computer SAU

Oracle VM (Ubuntu 14.04) PSAU

PostgresSQL

Octave

State estimatorWrite SE results

IDE4L standard RDBMS

Read measurementsRead grid data

Oracle VM (Ubuntu 14.04) SSAU

MMS Client Write SE at SS

Rep

ort w

ith SE at SS

Digital connection

PostgresSQL MMS Server


SCL file server

IEDRTDSMV IEDRTDSMVgen

Oracle VM (Ubuntu 14.04) VIEDs

PostgresSQL MMS Server


SCL file server

IEDRTDS IEDRTDSgen

MMS Report

MMS write

MMS Server

SCL file server

Read SE at SS

MMS Client

MMS Report

MMS write

Write measurements

Read set points

Write SE at SSWrite measurements

Read set points

Power Control

Read grid status and forecast

Write control set points

Python

ForecasterRead historian

Read weather forecastWrite power forecast

PostgresSQL

Octave

State estimatorWrite SE results


Read measurementsRead grid data

Power Control

Read grid status and forecast

Write control set points

Python

ForecasterRead historian

Read weather forecastWrite power forecast

DLMS/COSEM client

DLMS/COSEM readWrite

measurements

C37.118 client

Write measurements

C37.118 reports

• PMUs delivery of synchrophasors of current and

voltage. • Smart Meter

delivery of power, voltage and energy data• 4 virtual IEDs

MMS server developed by RWTH installed IDE4L PostgresSQL database for storing of data

• PSAU State estimation and power control developed

by TUT installed in Octave environment Load and generation forecasters developed by

UC3M installed in Python environment MMS Client developed by Unareti DLMS/COSEM client developed by Unareti OpenPDC client configured by RWTH Postgres SQL Database MySQL Database (for PMU data)

• SSAU State estimation and power control developed

by TUT installed in Octave environment Load and generation forecasters developed by

UC3M installed in Python environment MMS Client developed by Unareti MMS Server developed by RWTH DLMS/COSEM client developed by Unareti OpenPDC client configured by RWTH Postgres SQL Database


RMS voltage time profiles forLV nodes in Mid-Season weekend day afternoon scenario

Active Power time profiles for LV nodes in Mid-Season weekend day afternoon scenario

Reactive Power time profiles for LV nodes in Mid-Season weekend day afternoon scenario


ideal grid for all

Unareti real network demonstration


Secondary substation involved in the LV test

phase

ideal grid for all


• No voltage violations as expected


ideal grid for all


• Reactive power is controlled to minimize losses


ideal grid for all

Conclusions on secondary control demonstrations• The algorithms (load and production forecaster, state

estimator, secondary power control) operated correctly together

• The secondary control was able to mitigate congestions and to optimize network state

• Graded time operation of AVC relays is usually adequate to prevent adverse interactions between LV and MV secondary control

• The implemented optimization algorithm was in some cases too slow commercial solver should be used


ideal grid for all

Tertiary control


ideal grid for all

Tertiary control operates only at MV level


PL1PL2

PS0023

Load type Power factor Fixed Flexible

Domestic 0.9 59% 41%

Non-domestic (LV) 0.9 53% 47%

MV load 0.95 53% 47%

Type of flexibility Flexibility price (€/kWh)

Domestic consumers 0.15

Non-domestic consumers 0.12*

MV consumers 0.09*Obtained by interpolation.

Assumptions in the market agent

calculations:

ideal grid for all

Operation of network reconfiguration algorithm

• The first section of feeder 1 between nodes PS0023 and 545 is loaded at 102.5 % of its nominal capacity.

• Network reconfiguration algorithm operates and shifts part of the load of feeder 1 to feeder 2 by opening the breaker 468-827 and closing the breaker 603-117


Branch loading 102.5%

After reconfiguration

ideal grid for all

Operation of the market agent algorithm

• Line section at the beginning of feeder 3 between nodes PL2 and 1056 is loaded at 104.24 % of its nominal capacity

• Market agent algorithm purchases flexibility products to mitigate the congestion


Branch loading 104.24%

Node Flexibility (kW)

60 26.4

297 15.1

585 3.7

604 6.1

732 6.2

749 11.1

870 7.2

937 1.2

987 39.2

1056 13.7

1070 18.6

1136 31.2

1143 7.8

1180 21.3

1187 3.9

1190 1.5

1462 6.2

2236 0.9

ideal grid for all

Also economical evaluation is needed

• Demonstrations verify the correct operation of the implemented automation architecture and algorithms but do not tell anything on the cost-efficiency of alternative methods also network planning methods have to be developed

• Congestion management affects both CAPEX and OPEX• Yearly operational costs of alternative methods can be calculated

using hourly load flow calculations (losses, cost of curtailed generation etc.)

• Investment costs of alternative methods consist of network reinforcement costs and/or costs of automation system

• The network total costs can be compared by combining the investment cost annuities and yearly operational costs






• Demonstrations

• Conclusions

ideal grid for all

Results

SLIDE 83

1. Active network management concept

2. Hierarchical and decentralized automation architecture

3. Distribution grid and DER management functionalities

4. Benefits and impacts of functionalities for DSO

5. Demonstrations of concept, architecture and functionalities in three field demonstrations

6. Recommendations and roadmap


ideal grid for all

Substation Automation Unit


SAU

Interfaces

Database

Applications

DLMS/COSEM

MMS

Prosumers – Smart meters and DER IEDs

Control Center - DMS

Substations - IEDs

IEC 61850

CIMManagement

model

Bridge model

State Estimation

Power control

ForecastData Acquisition

Reports

ideal grid for all

Conclusions

• Basis for distributed grid management and interaction of business players

• Design, implementation and demonstration of ANM, hierarchical and distributed automation architecture and commercial aggregator concepts

• Efficient utilization of grid assets• Monitoring and control of complete grid

• Increased hosting capacity for RESs and DERs

• Enhanced reliability of power supply

• Planning tools estimate the hosting capacity increment


ideal grid for all

Conclusions

• Scalability of automation solution• Automation is based on existing devices allows to

gradually deploy the new solutions

• The same architecture and cores of the automation are suitable for both primary and secondary substations

• Functions can be deployed locally and coordinated

• Local functions are light, makes integration scalable

• Vertical and horizontal integration provides a complete view of the distribution network status

• Data exchange between DSO and aggregator to validate, purchasing and activating flexibilities will further extend ANM capabilities


ideal grid for all

Conclusions

• Utilization and development of standards • IEC61850, DLMS/COSEM (IEC 62056) and CIM (IEC

61970/61968) for architecture and implementation

• The interoperability has been achieved through:• Identification of interfaces and information exchange

synthesized from the IDE4L use cases.

• The automation architecture was hence derived and defined based on the SGAM framework.

• Implementations of automation system have been demonstrated in three demonstration sites


ideal grid for all

Thank you!

www.ide4l.eu [email protected]

[email protected]

http://www.ide4l.eu/

mailto:[email protected]

mailto:[email protected]

distributed and hierarchical congestion management in ... · for monitoring & control systems...

Documents