multiDC: Controlling the Power Flows · if f < threshold, then transfer = xx MW • Recommendation: move from stepwise-triggers to droop-frequency control – Sharing stops as soon

Spyros Chatzivasileiadis, DTU

multiDC: Controlling the Power Flows

Three main drivers

• 100% renewables

• 100% inverter-connected devices

• HVDC Lines and Grids

Spyros Chatzivasileiadis - multiDC: Controlling the power flows 219 September 2019

Denmark is unique in being involved in two first-of-their-kind projects

Kriegers Flak: First interconnection in the world to integrate off-shore wind

North Sea Energy Hub: First hub-and-spoke topology for offshore wind


multiDC

Innovative Methods for Optimal Operation of Multiple HVDC Connections and Grids

• Innovation Fund Denmark Grand Solutions

• Partners:

– Two neighboring TSOs: Energinet, Svenska kraftnät

– Three universities: DTU, KTH, Univ. of Liege

– One major manufacturer: ABB

– Advisory Board: RTE, Nordic RSC

• 25.7 million DKK (3.5 million EUR)

• 4 years; Start May 1, 2017, ends in 2021


The Team

Spyros Chatzivasileiadis - multiDC: Controlling the power flows 5

18 persons – 3 countries – 10 nationalities

19 September 2019

Question #1:What is the optimal topology for the

North Sea Energy Hub?


AC

AC

UK

NetherlandsGermany

Denmark

Norway

AC Ring on the island

AC50 Hz

Low-frequency AC (16.67 Hz)

HVDC Point-to-point HVDC


ACAC

UK

NetherlandsGermany

Denmark

Norway

DC Ring on the island

VSC

HVDC Multi-terminal HVDC Grid


For now: Focus on the AC ringChallenges and Opportunities

• Zero-inertia AC Ring– Fast transients

• How to guarantee N-1 security?

– Coordination of VSC converters

• Sharing wind power among the countries– Ownership of wind farms

– Do we need to adapt the market structures?


Low Frequency AC has a larger stability region and allows longer distances, but costs and weight of transformers may cancel out the benefits

• 16.67 Hz leads to larger stability region than 50 Hz

But:

• The costs for 16.67Hz transformers are 3x higher

• The weight for 16.67Hz transformers is 3x higher

50 Hz 16.67 Hz


Zero-inertia Low-inertia


Zero-inertia Low-inertiaGrid-forming

Grid-following


Zero-inertia vs. low-inertia configuration

1. Which topology can better withstand disturbances?

2. How should the converters share the control effort to keep the frequency within limits on the island?



1. Which topology can better withstand disturbances?

2. How should the converters share the control effort to keep the frequency within limits on the island?

Answers in the next talk by Georgios Misyris:

North Sea Wind Power Hub: What is the optimal topology?



1. What is the response of the low-inertiaconfiguration under different disturbances?


Demo by Matas Dijokas:

RMS vs EMT: The need for new simulation tools

Demo by Thierry van Cutsem/Bertrand Bastin:

The transition to zero-inertia systems: synchronous condensers in the North Sea Energy Hub

2. What is the response of the zero-inertia configuration under different disturbances?

The RMS model does not capture all unstable modes

RMS simulations are not enough for zero- and low-inertia systems

• The dynamics for zero/low inertia systems are too fast

• RMS simulations, such as in Powerfactory, cannot capture all instabilities

• Need for EMT; but EMT are very computationally intensive

Unstable regionStable region


The RMS model does not capture all unstable modes

RMS simulations are not enough for zero- and low-inertia systems

Zero-inertia systems

1. For what type of disturbances is RMS still good?

2. For what phenomena do we need new simulation tools?

Unstable regionStable region

Demo by Matas Dijokas:

RMS vs EMT: The need for new simulation tools


Question #2:Can we save costs from

better HVDC coordination?


During low-inertia periods, we cannot afford the loss of large generation units

• Low-inertia: lower frequency nadir during disturbances

• N-1 Security: Operators cannot afford the loss of large generation units during low-inertia periods

• Need for redispatch:

– Reduce generation of large units

– Increase generation of smaller units

1919 September 2019 Spyros Chatzivasileiadis - multiDC: Controlling the power flows

Oskarshamn 3

Cost: 1.85 million Euros• Happened 3 times in Sweden

during Summer 2018

Use HVDC to support frequency

• Recommendation: Instead of internal redispatching, share reserves over HVDC

– In case of fault, use HVDC Emergency Power Control and reserves from neighboring countries

• Cost savings of up to 14 million Euros/year by 2025


Oskarshamn 3

Balticcable

More answers in the talk by Andrea Tosatto:

Cost Savings from HVDC coordination

How can we reduce the costs even more?

Focus on the Emergency Power Control (EPC) of HVDC

Current scheme for sharing reserves over HVDC

• Stepwise-triggers:

if f < threshold, then transfer = xx MW

• Recommendation: move from stepwise-triggers to droop-frequency control– Sharing stops as soon as frequency starts to recover

– Lower power and energy exchange


Demo by Danilo Obradovic:

Advanced Control for Sharing Reserves over HVDC

Zero-price difference between zones does not cover HVDC losses

22Spyros Chatzivasileiadis - multiDC: Controlling the power flows

2017 2018

% LOSSES % LOSSES

KONTISKAN 61% 1.2 M€ 53% 1.5 M€

STOREBÆLT 73% 0.8 M€ 74% 1.1 M€

SKAGERRAK 47% 3.2 M€ 46% 4.7 M€

ESTLINK 76% 3.1 M€ 95% 6.7 M€

FENNOSKAN 99% 3.8 M€ 80% 4.2 M€

12 M€ 18 M€

Hours of operation with zero-price-difference and corresponding losses (€)

Source: https://www.nordpoolgroup.com/

19 September 2019

Introduction of linear HVDC loss factors


EE

FI

SE3

SE2

SE1

NO1

NO2

NO5NO3

NO4

* Fingrid, Energinet, Statnett, Svenska Kraftnät, Analyses on the effects of implementing implicit grid losses in the Nordic CCR, Tech. Rep., April 2018

• Nordic TSOs, April 2018: Analyses onthe effects of implementing implicitgrid losses in the Nordic CCR*

19 September 2019

HVDC loss factors: Key insights

1. HVDC loss factors can unfairly penalize HVDC lines and could lead to higher costs

– Developed methodology to consider both AC and HVDC losses

2. Losses are non-linear. Linear loss factors unfairly penalize one HVDC line over the other

– Can affect merchant HVDC lines

– Recommendation: use piece-wise linear loss functions

More answers in the talk by Andrea Tosatto:

Cost Savings from HVDC coordination


Open-source models

25Spyros Chatzivasileiadis - multiDC: Controlling the power flows19 September 2019

• 424 buses

• 80 synchronous machines

• 2 asynchronous areas

• Open source (including HVDC models)

• Modeling both in RMS and EMT

• Powerfactory and RAMSES (U.Liege)

Nordic dynamic model

26Spyros Chatzivasileiadis - multiDC: Controlling the power flows19 September 2019

Kriegers Flak: First interconnection in the world to integrate off-shore wind

Live Demo of Kriegers Flak Master Controller


Live Demo by Rüdiger Franke:

Live Demonstration of the KriegersFlak controller

19 September 2019

Nordic market model

• Both for zonal and nodal markets

• Grid reduction for flow-based market coupling

− Estimation of the equivalent PTDF matrix

PL

LT

EE

RU

FRE

FI

SE4

SE3

SE2

SE1

NO1

NO2

NO5

NORDIC EQUIVALENTDenmarkNorwaySwedenFinlandNeighboring countries

AC interconnectionsHVDC interconnections

RU

NLDE

PL

LT

EE

NO4

NO3

NO5 NO1

NO2

SE1

SE2

SE3

SE4

DK2

FI

DK1


Planned Next Steps

• Investigation of DC ring topology for NSEH

• HVDC corrective control as a market product

• Implement and Demonstrate at PowerlabDK– RTDS: Low inertia vs Zero inertia NSEH

• Use of a real controller for the Synchronous Condenser

– Impact of RMS vs EMTP

• Hopefully test the most promising functions atEnerginet!


Next Steps for this Event

AGORA (here)1. Advanced control for sharing of

reserves over HVDC

2. RMS vs EMT: the need for new simulation tools


PowerlabDK Control Room3. Live Demonstration of the Kriegers Flak

controller

4. The transition to zero-inertia system: synchronous condensers in the North Sea Energy Hub

• Cost Savings from HVDC coordination

Two presentations (multiDC in detail)• North Sea Wind Power Hub – What is the optimal

topology?

Four Demonstrations Split in two areas (2 groups)

20 min. each, repeated for each group

19 September 2019

Panel Session

Peter Lundberg

ABB

Product Manager HVDC Light


Lykke Mulvad Jeppesen

Ørsted

Head of Regulatory Affairs, Energy

Economics

Göran Ericsson

Svenska Kraftnät

Head of R&D

Lena Kitzing

DTU

Head of Energy Economics and

Regulation

Fitim Kryezi

Energinet

Technical Project Manager

“The role of interconnections and international collaboration”

19 September 2019

Moderator:

Jacob Østergaard

DTU

Professor, Head ofCenter

Thank [email protected]

www.multi-dc.eu

Documents

multiDC: Controlling the Power Flows · if f < threshold, then transfer = xx MW • Recommendation: move from stepwise-triggers to droop-frequency control – Sharing stops as soon