Participation in ancillary service and physical electricity markets with flexible demand resources

67 Pages·2017·2.42 MB·English

by  Mirja Ruotsala

Santeri Oksanen

Participation in ancillary service and physical electricity markets
with flexible demand resources

Thesis submitted in partial fulfilment of the requirements for
the degree of Master of Science in Technology.

Helsinki, May 26th, 2017
Supervisor: Professor Sanna Syri
Advisor: Outi Matikainen, M.Sc. (Tech.)
Aalto-yliopisto, PL 11000, 00076 AALTO
www.aalto.fi

Diplomityön tiivistelmä

Tekijä Santeri Oksanen
Työn nimi Participation in ancillary service and physical electricity markets with flexi-
ble demand resources
Koulutusohjelma Energia- ja LVI-tekniikan koulutusohjelma
Pää-/sivuaine Energiatekniikka Koodi Ene
Työn valvoja Professori Sanna Syri
Työn ohjaaja(t) Diplomi-insinööri Outi Matikainen
Päivämäärä 26.5.2017 Sivumäärä 63 Kieli Englanti
Tiivistelmä

Diplomityön tavoitteena oli selvittää keskisuurten sähkökuormien aggregointi- ja
kysyntäjoustomahdollisuuksia sähköverkon tasapainottamiseksi. Fokus työssä oli
taloudellisen kannattavuuden arvioimisessa. Työssä arvioitiin myös Fingridin ja Seam
Oy:n välisen aggregointipilotin suorituskykyä taajuuden hallinnan näkökulmasta.

Taloudellisten hyötyjen selvittämiseksi tutkittiin aluksi sekä Pohjoismaisia fyysisiä
sähkömarkkinoita, että Fingridin ylläpitämiä reservimarkkinoita. Markkinapaikkojen
osalta arvioitiin sinne osallistuvilta kuormilta vaadittavia ominaisuuksia.

Aggregointipilotista havaittiin rebound-ilmiön olevan merkittävä. Ilmiötä havaittiin
kuitenkin voitavan pienentää laitosten ohjaukseen tehtävin pienehköin muutoksin.

Lopuksi työssä kehitettiin työkalu joustoon kykenevän kuorman säästöpotentiaalin
arvioimiseksi kysyntäjoustoratkaisuja käyttöön otettaessa. Työkalua esiteltiin
optimoimalla pakkasvarastolta saatuja sähkön kulutustietoja, eri markkinoille
osallistuessa. Tuloksista havaittiin spot-optimoinnin lisäksi kapasiteetin myymisen
taajuudenhallintareserveihin tuovan selkeimmät säätöt laitoksen sähkön
hankintakustannuksiin.

Avainsanat Kysyntäjousto, sähkömarkkinat, aggregointi, reservimarkkinat,
taajuussäätö, spot-optimointi
Author Santeri Oksanen
Title of thesis Participation in ancillary service and physical electricity markets with
flexible demand resources
Degree programme Energy Technologies
Major/minor Energy technology Code Ene
Thesis supervisor Professor Sanna Syri
Thesis advisor(s) Outi Matikainen, M.Sc. (Tech.)
Date 26.5.2017 Number of pages 63 Language English
Abstract

The purpose of this thesis was to investigate the possibilities for demand response and
aggregation for medium sized electric loads to balance the electric grid. The focus was on
estimating the achievable economic benefits. The thesis also evaluated the performance
of a pilot project on aggregating resources for frequency regulation reserves.

To evaluate the economic benefits the physical and ancillary electricity markets in the
Nordic countries was analyzed first. For each marketplace, general requirements for a
load to participate was presented.

In the pilot project, the rebound-effect was seen significant. It was however also noted,
that it is possible to reduce the effect by relatively small changes in the facilities control
principles.

Finally, a tool was developed to evaluate the economic benefits achievable by introducing
demand response activities to a flexible load. The tool was demonstrated by optimizing
the electricity consumption of a cold storage facility when participating in different mar-
kets. From the results, it can be seen that the combination of spot optimization and par-
ticipation in the frequency controlled reserves are the most lucrative scenarios.

Keywords Demand response, electricity markets, aggregation, ancillary service, fre-
quency regulation, spot-optimization
Acknowledgements

This master’s thesis has been written for Seam Oy between fall of 2016 and the spring of
2017. The work discusses the benefits of introducing demand response in many forms for
customers with flexible, controllable loads.

Mostly, I would like to thank Jukka Jaatinen at Seam Oy for providing me with this very
interesting and actual topic and the support while writing it. It was very interesting to
study the field, as there is a lot of ongoing research and piloting. I would also like to
thank my advisor, Outi Matikainen, very much, as she guided me during the process and
insight about the Nordic electricity markets. Laura Ihamäki at Fingrid gave me valuable
information about the ancillary markets operated by Fingrid. It was also very interesting
to meet with Seam’s customers.

Finally, I would like to express my greatest gratitude to my supervisor, professor Sanna
Syri, who encouraged me to speed up with the thesis to finish my thesis prior to the sum-
mer holiday season.

In Helsinki May 26th, 2017

Santeri Oksanen
Table of Contents

Tiivistelmä
Abstract
Acknowledgements
Table of Contents .............................................................................................................. 1
Used symbols .................................................................................................................... 3
List of abbreviations .......................................................................................................... 4
1 Introduction ............................................................................................................... 5
1.1 Power balance in the electric grid ...................................................................... 5
1.2 Demand response ............................................................................................... 6
1.3 Research questions ............................................................................................. 6
2 Electricity markets in the Nordic countries ............................................................... 8
2.1 Elspot-market ..................................................................................................... 9
2.2 Elbas ................................................................................................................. 11
2.3 Flexible loads in the physical electricity markets ............................................ 12
2.4 Summary .......................................................................................................... 12
3 Ancillary services .................................................................................................... 13
3.1 The frequency containment process ................................................................. 13
3.2 Frequency containment reserves ...................................................................... 14
3.2.1 Market size ................................................................................................ 15
3.2.2 Normal frequency containment reserves (FCR-N) ................................... 17
3.2.3 Disturbance frequency containment reserves (FCR-D) ............................ 20
3.2.4 Activation of FCR-D relay decoupled loads ............................................. 21
3.3 Reporting of FCR-capacity availability ........................................................... 22
3.4 Frequency restoration reserves (FRR) .............................................................. 23
3.4.1 Automatic frequency restoration reserves (aFRR) .................................... 23
3.4.2 Balancing power market ........................................................................... 25
3.4.3 Quick disturbance reserves ....................................................................... 27
3.4.4 Balancing power capacity markets ........................................................... 27
3.5 Balance settlement ............................................................................................ 28
3.6 Summary .......................................................................................................... 29
4 Case study ............................................................................................................... 31
4.1 Similar studies .................................................................................................. 31
4.2 Seam ................................................................................................................. 32
4.3 Participating resources ..................................................................................... 33
4.3.1 Heat pump ................................................................................................. 33
4.3.2 Cold storage .............................................................................................. 33
4.3.3 Small hydropower plant ............................................................................ 33
4.4 Rebound effect ................................................................................................. 34
4.4.1 Reasons for rebound effect........................................................................ 35
4.4.2 Possibilities to reduce the rebound effect ................................................. 38
4.5 Economic feasibility ......................................................................................... 38
4.5.1 Cold storage facility .................................................................................. 38
4.5.2 Hydropower plant...................................................................................... 39
4.6 Experience gained from the pilot ..................................................................... 40
5 Suitable loads for demand response ........................................................................ 42
2

5.1 General requirements ....................................................................................... 42
5.1.1 Storage capacity ........................................................................................ 42
5.1.2 Measurement ............................................................................................. 43
5.1.3 Controllability ........................................................................................... 43
5.1.4 Process dependencies ................................................................................ 44
5.2 Available control capacity ................................................................................ 44
5.3 Potential for demand response in the industry in Finland ................................ 45
6 Simulating a flexible load ....................................................................................... 47
6.1 Common assumptions ...................................................................................... 47
6.2 Simulator properties ......................................................................................... 47
6.2.1 Constraints ................................................................................................ 48
6.2.2 Cost function ............................................................................................. 48
6.3 Scenarios .......................................................................................................... 49
6.4 Optimized load and properties ......................................................................... 50
6.5 Examples of optimized consumption of electricity .......................................... 51
6.5.1 Consumption on January 7th ...................................................................... 51
6.5.2 Consumption on July 26th ......................................................................... 52
6.6 Results .............................................................................................................. 53
6.7 Suggested demand response strategy ............................................................... 54
6.7.1 Frequency controller driven load .............................................................. 54
6.7.2 On/Off controllable load ........................................................................... 55
6.8 Sources of uncertainty ...................................................................................... 56
6.9 Summary .......................................................................................................... 57
7 Conclusions ............................................................................................................. 58
3

Used symbols

C Capacity
CR Capacity requirement
E Energy
f Frequency
H Inertia
J Moment of inertia
P Power
R Reserve capacity
t Time
T Temperature
4

List of abbreviations

aFRR Automatic Frequency Restoration Reserves
AC Alternating Current
COBYLA Constrained Optimization by Linear Approximation
COP Coefficient of Performance
DC Direct Current
DI Dimensioning Incident
DR Demand Response
EV Electric Vehicle
FCR Frequency Controlled Reserve
FCR-N Frequency Controlled Reserve, Normal
FCR-D Frequency Controlled Reserve, Disturbance
FRR Frequency Restoration Reserve
FRR-M Frequency Restoration Reserve, Manual
ICT Information and Communication Technology
TSO Transmission System Operator
5

1 Introduction
To keep the lights on at homes and city streets, factories running and refrigerators cold,
the consumption and generation of electricity must match each other at every moment.
Due to the fluctuating nature of electricity consumption, this might be a challenging task.

Traditionally electricity has been produced mainly by large centralized power plants run-
ning on for example coal, nuclear, gas or hydro power. The power output of these sources
of electricity are controllable by nature, and thus the natural solution has been to adjust
the output of power plants to match the demand of electricity. (International Energy
Agency, 2016)

Due to raising concerns about climate change – mainly caused by the combustion of fossil
fuels – there has been growing interest and actions to replace the traditional fossil fuel
based electricity generation with new renewable sources, such as wind and solar power.
Many of these new sources of electricity are intermittent by nature, and as such they are
hard – if not impossible – to adjust to match with the demand of electricity. This raises
the transmission system operators (TSO’s), who are responsible for the balance of power
generation, with new challenges of how to keep the balance between power generation
and consumption. (IPCC, 2014)

One obvious part-solution is to adjust demand to match with generation. In the past, this
has been hard to accomplish, but the rapid development in the Information Technology
and Telecommunications (ITC)-sector during the last decades has not only lowered the
price of programmable logics and data communications, but also increased their reliabil-
ity significantly. This together with the rise of renewable power generation has led in
growing interest towards demand response (DR) solutions.
1.1 Power balance in the electric grid
The frequency of the Alternating Current (AC) electric grid is typically – depending on
the market – rated either at 50 or 60 Hz. Any small mismatch between the generation and
consumption of electricity is seen as small fluctuation in the actual frequency. If there is
too little generation, the frequency drops below the rated value, and vice versa if there is
too much generation (Fingrid, 2016b).

Traditional power generation is based on large turbines and synchronous generators,
which means that the rotation speed of the turbine and the generator is always an exact
multiple of the grid frequency. This causes the turbine and the generator to slow down if
the frequency of the grid is decreasing, and vice versa when the frequency is increasing.
The masses of the turbine and generator are heavy, and thus the inertia of the combination
is high. High inertia leads to natural resistance towards changes in the frequency, as the
decrease in frequency leads to the turbine and generator to release extra energy when
decelerating, or absorb extra energy when accelerating in the occasion of the frequency
increasing, which balances the grid. The inertia (H) of a generator is calculated according
to the following formula, where the kinetic energy stored in the rotating turbine-generator
is divided by the rated power of the generator. The result represents how many seconds
the generator could provide power without power input from the turbine.

𝐻 = 𝐸𝑘𝑖𝑛 = 𝐽(2𝜋𝑓𝑚)2 (1)
𝑃𝑅 2𝑃𝑅
6

New renewable energy generation based on wind or solar power does not have inbuilt
inertia, as the produced electricity is converted to AC via an inverter, which is not featur-
ing rotating masses. Thus, the grid frequency gets more volatile as the share of these
renewables grows. In particular, the frequency drop caused by a drop of a large generating
unit grows significantly (Biegel, Westenholz et al. 2014).

If the difference between the rated and the actual frequency grows too much, the appli-
ances connected to the electric grid may be damaged. There is also a high chance of a
blackout. Due to these reasons, the TSO’s have defined an allowed band in which between
the frequency may alter. To keep the frequency within the allowed limits, the TSO’s are
purchasing balancing reserves that are activated according to grid frequency.
1.2 Demand response
The rise of the ICT-sector during the last decades has revolutionized how people com-
municate with each other and how they search for information. While the driving forces
of the development have been in peoples need to communicate with each other and search
for information, it has also led to a reliable and affordable data communication possibili-
ties available worldwide. Additionally, the rapid development in power electronics has
speeded up the deployment of new innovations such as frequency converters, which al-
lows the processes that utilize this kind of controls to tune their energy consumption ex-
actly. Taking these two factors into account, adjusting demand according to external fac-
tors – such as power balance – has become both technically and economically feasible.

Demand response may be divided to multiple categories:
- Load shifting
- Load following
- Peak shaving

As the name indicates, load shifting indicates a process where load is shifted from expen-
sive to cheaper hours. Cold storage facilities are an example of such consumers which
may benefit from this service, as the storage facility may be chilled more when the elec-
tricity is more affordable, and in return let the temperature rise during expensive hours.
Generally, all consumption with inbuilt storage capabilities are a good fit for load shifting.
(Siano, 2013; Jaatinen, 2016)

Load following on the other hand translates to consumption following the momentary
load balance more accurately, adjusting its consumption based on for example grid fre-
quency. As with load shifting, this kind of controls works perfectly with consumption that
has inbuilt storage capabilities, but the required timeframe is shorter, as the consumption
is moved for a shorter time. (Jaatinen, 2016)

The simple reduction of consumption during peak-hours translates to peak shaving. Sep-
arating it from load shifting is not simple, and basically all load shifting means peak shav-
ing at the same time. Peak shaving does not however require shifting the consumption to
a later time. An example could be a swimming pool heating, which is turned off during
peak consumption hours. The temperature of the pool is reduced during these hours, but
it may be acceptable due to reductions in the heating bill. (Siano, 2013; Jaatinen, 2016)
1.3 Research questions
This thesis starts by presenting both the physical and ancillary electricity markets in the
Nordic countries. The goal of this part is to define the marketplaces that are the most