SEAI sponsored report on Electricity Storage Technologies, May 2004

SEAI sponsored report on Electricity Storage Technologies, May 2004
UCC Sustainable Energy Research Group
FINAL REPORT
Project - RE/HC/03/001
Funded by
1 of 1
Study of Electricity Storage Technologies and
Their Potential to Address Wind Energy
Intermittency in Ireland
Final Report
prepared by
Dr. Adolfo Gonzalez
Dr. Brian Ó Gallachóir
Dr. Eamon McKeogh
Sustainable Energy Research Group,
Department of Civil and Environmental Engineering,
University College Cork
and
Kevin Lynch
Rockmount Capital Partners,
Cork
May 2004
Funded by the National Development Plan through Sustainable Energy Ireland’s
Renewable Energy Research, Development and Demonstration Grant RE/HC/03/001.
Electricity Storage and Wind Energy Intermittency
Final Report
EXECUTIVE SUMMARY
Context
Wind energy deployment in Ireland is due to accelerate very rapidly in the short term.
Until recently, the rate of wind energy deployment in Ireland has been relatively
modest at approximately 20MW installed capacity, on average per year, since 1997.
The deployment rate increased since 2003 and the total amount of wind power
installed by May 2004 was 190MW, with a further 44MW nearing final connection
(CER 2004).
Adding to this the amount of wind power not yet installed but with signed connection
agreements raises the 190MW to 823MW, with a further 749MW within the process
and an additional 627MW for which applications were being checked in May 2004.
This represents a cumulative total of 2199MW, although it is unclear how many of
these wind farms have secured planning permission and the necessary finance to
enable construction. A survey carried out in 2004 (CER 2004) concluded that 661MW
would be likely to connect to the system by mid 2006, bringing the total installed
wind capacity to 851MW. This represents 14% of projected total generation capacity
in 2006 (ESBNG 2003).
Despite the recent dramatic increase in activity, it is almost certain that the target in
the Green Paper on Sustainable Energy will not be met. The target set was for an
additional 500MW of renewable generated electricity to be delivered by 2005, i.e.
before January 1 2005. Excluding AER III wind farms that fell outside this target,
92MW has been delivered to date and based on the timeframe envisaged for
connections (ESBNG 2003), an additional 308MW will be connected by the end of
2004, delivering thus 400MW of the 500MW total. Based again on accepted
connection offers, it is envisaged that an additional 186MW of wind energy capacity
will be installed during 2005. This assumes that the wind farms with connection
agreements will be backed financially and will not encounter unforeseen
circumstances that might delay or even prevent completion.
One of the difficulties in accommodating wind energy intermittency in Ireland relates
to the nature of the electricity network. The design of the network evolved along the
conventional approach, where large scale thermal plant feed into a transmission
network through to a distribution network and supply to final customers. The structure
and operation of the network does not readily accommodate decentralised embedded
generation such as wind farms. This problem is common to other European electricity
networks but Ireland faces an additional challenge in accommodating wind energy
due to the lack of adequate scale and interconnection.
As electricity demand has grown in the past decade, the transmissions system has
become strained, prompting the urgent need for system upgrade. A major
refurbishment and expansion programme running from 2001 to 2005 is underway.
This programme increases annual capital expenditure on the transmissions and
distribution networks by a factor of three, contributing to the recent electricity price
increases. Over €2.6 billion is being invested in the high voltage and low voltage
networks, particularly in the counties along the southern and western coasts. Over
1
Electricity Storage and Wind Energy Intermittency
Final Report
€820 million will be spent on transmission, over €1 billion on distribution renewal
(including conversion of 50% the 10 kV network to 20 kV), and over €665 million on
distribution reinforcement.
This upgrading of the electricity networks will improve the system and thereby
facilitate the accommodation of increased wind energy penetration. However, this
upgrading programme should now be reviewed to take account of the anticipated
accelerated deployment of wind energy outlined above in line with the
recommendation of the Renewable Energy Strategy Group (2000). This should be
done within the framework of an integrated sustainable energy policy for Ireland.
This is separate from the financial mechanism for grid connecting wind farm clusters,
which has been examined by the Steering Group for Grid Upgrade Development
Programme and adopted by CER.
ESB National Grid, the Irish Transmission System Operator has expressed a number
of concerns at the amount of wind energy projects currently possessing (together with
those seeking) signed connection agreements. These concerns relate primarily to the
inability to ride through faults and the impacts on the stability of the system. In the
absence of dynamic models for wind turbines intending to connect, the nature of some
of these impacts is unclear.
There are other concerns (associated with reserve requirements) that relate to the
intermittency of wind energy and the high level of variability in the output of each
wind farm. The primary focus of this study is the role of energy storage in addressing
the these concerns.
There is much debate as to how much intermittent wind energy an isolated grid like
Ireland’s can support without impacting negatively on system security and stability.
For ESB National Grid, penetration levels of 5 – 7% are acceptable compared with
views expressed by industry of over 30%, the latter under assumptions of adequate
interconnection and well developed enabling technologies such as dynamic load
levelling and short term wind forecasting.
In addition to the impacts on penetration levels of wind, intermittency also has a key
bearing on the value placed on wind generated electricity. This is particularly
apparent in liberalised electricity markets, where wind farm developers are not
guaranteed a fixed feed in tariffs for the output of their plant (as is the case for
example in Germany and Denmark). The degree to which intermittency affects the
financial value of the electricity will depend to a large extent on the trading rules that
apply in each electricity market. The proposal by CER, for example, that all
generators pay for the cost of reserves in line with a ‘causer pays’ principle will
typically mean a higher charge for wind generators due to the additional reserve
requirements attributable to intermittency.
Electricity storage
Energy storage is not a new concept in the electricity sector. Utilities across the world
have built a number of pumped-hydro facilities in the last few decades, resulting in a
2
Final Report
Electricity Storage and Wind Energy Intermittency
storage component of roughly 5% the capacity of all the European countries, 3% in
the US, and 10% in Japan.
These pumped-hydro plants, and to a lesser extent compressed air storage systems,
have been used for load levelling, frequency response, and voltage control. Likewise,
storage facilities based on other technologies such as lead-acid batteries have been
installed by a number of utilities to fulfil a variety of functions. At a different scale,
energy storage is also commonly used at the user level to ensure reliability and power
quality to customers with sensitive equipment. Another traditional application is the
electrification of off-grid networks and remote telecommunications stations, mostly in
connection with renewable sources.
The applications for electricity storage technologies can be grouped as:•
load management (load levelling, ramping and load following);
•
spinning reserve (fast response and conventional);
•
system stability and voltage regulation;
•
deferral of system and plant upgrading;
•
renewable energy applications;
•
end use applications (UPS, peak shaving and emergency back-up).
Traditionally, electricity storage technologies have been used for the technical
benefits they bring to electrical systems. With the arrival of liberalised electricity
markets, a new application for energy storage has presented as price arbitrage (buying
at low price, storing and selling at a high price). The key characteristics of storage
technologies that determine which applications they are most suited for are :
•
discharge duration;
•
power rating;
•
energy storage capacity;
•
response time
•
costs in the context of benefits.
Electricity storage systems can be categorized as mechanical, electromagnetic and
electrochemical storage devices as follows.
Mechanical
Electromagnetic
Electrochemical
Pumped Hydro
Super-Capacitors
Batteries
Compressed Air
Super-Conducting
Magnets
Flow Batteries
Flywheel
Hydrogen
3
Electricity Storage and Wind Energy Intermittency
Final Report
Wind intermittency and storage
Many of the problems associated with wind energy intermittency can be addressed
using appropriate electricity storage technologies. Different problems are associated
with
•
short duration fluctuations (seconds) – leading to power quality problems;
•
hourly variations – primarily market related (including payment for reserve);
•
longer term variations (days) –affecting the requirements for backup and the
‘firmness of wind energy’
The key electrical storage technologies that are appropriate for each of these problems
are :Power Quality
flywheels
hydrogen
batteries
flow batteries
Market Related
pumped hydro
hydrogen
batteries
flow batteries
compressed air
Long term fluctuations
pumped hydro
hydrogen
batteries
flow batteries
compressed air
It is important to note that these technologies are at varying stages of technological
maturity.
•
Pumped hydro energy storage (PHES) is a mature and familiar technology and
has been utilised within electricity systems for many years. It is the most
widespread energy storage system currently in use on power networks, operating
at power rating up to 4,000 MW and capacities up to 15 GWh. PHES uses the
potential energy of water, transferred by pumps (charging mode) and turbines
(discharge mode) between two reservoirs located at different altitudes. Currently,
the overall efficiency is in the 70-85% range although variable speed machines are
now being used to improve this. The efficiency is limited by the efficiency of the
deployed pumps and turbines (neglecting friction losses in pipes and water losses
due to evaporation). Plants are characterized by long construction times and high
capital costs. One of the major problems related to building new plants is of an
ecological/environmental nature.
•
Compressed air energy storage is also a mature technology but much less
deployed than pumped hydro. The electricity is stored by compressing air via
electrical compressors in huge storage facilities, mostly situated underground in
caverns created inside appropriate salt rocks, abandoned hard-rock mines, or
natural aquifers. Recovery takes place by expanding the compressed air through a
turbine, but the operating units worldwide incorporate combustion prior to turbine
expansion in order to increase the overall efficiency of the system. Hence CAES
can be regarded as peaking gas turbine power plants, but with a higher efficiency,
thanks to the decoupling of compressor and turbine, and much lower overall cost.
Deployment is often dependent on the availability of suitable underground
reservoirs but custom built high pressure storage tanks can be utilised.
•
Kinetic energy may also be used to store energy in the form of the inertia of a
flywheel. Flywheels have been used in hydro power stations with synchronous
4
Electricity Storage and Wind Energy Intermittency
Final Report
generators for many years. With the advent of advanced composite materials with
high tensile strength, and the development of stable magnetically suspended
bearings, flywheels may now be made with significantly higher operational
speeds. All reciprocating engines contain flywheels to smooth the pulsed output of
the pistons and provide stable power. Flywheels storage systems are particularly
suitable for power quality control. They can provide ride-through power for the
majority of power disturbances, such as voltage sags and surges, and can bridge
the gap between a power outage and the time required to switch to long-term
storage or generator power with excellent load following characteristics.
•
Capacitors store energy by way of separating the charge onto two facing plates.
They are widely used in electronic devices for power smoothing after rectifying.
Typically, these applications require very small energy amounts. In order to
increase the energy density, the so-called ‘Super-Capacitors’ (or even ‘Ultracapacitors’, if their capacitance exceeds 1000F) have been developed. They use
polarized liquid layers at the interface between a conducting ionic electrolyte and
a conducting electrode, which increases the capacitance. Super-Capacitors Energy
Storage (SCES) offers extremely fast charge and discharge capability, albeit with
a lower energy density than conventional batteries can provide and can be cycled
tens of thousands of times without degradation.
•
In a Superconducting Magnet Energy Storage (SMES) device, a coil of
superconducting wire allows a DC current to flow through it with virtually no
loss. The current creates a magnetic field that stores the energy. On discharge,
special switches tap the circulating current and release it to serve a load. To set the
coil in a superconducting state, it has to be cooled down either to 4.2°K (lowtemperature superconducting) or 77°K (high-temperature superconducting).
Technical improvements and a better knowledge of dealing with and controlling
cryogenic systems have allowed SMES to penetrate the market and compete with
more common storage systems. The dynamic performance of SMES is far
superior to most other storage technologies. Response times down to milliseconds
are possible and the energy can be transferred very quickly. SMES are most
suitable for high value/low energy applications, where the storage requirement is
for less than a few seconds, with power requirements up to 1 or 2 MW.
•
Batteries are the most common devices used for storing electrical energy.
Traditionally they have been used for small scale applications but there is growing
awareness amongst manufacturers of the potential applications for larger scale
energy storage in the context of liberalised electricity markets. As battery cells
have a characteristic operating voltage and maximum current capability, battery
systems normally consists of several cells, linked in line or parallel dependent on
the required power and energy rating. Batteries exhibit a fast response to changes
in power demand. Their efficiency varies among technologies, and also depends
on the application and the operation regime. The most mature technology, flooded
lead-acid (LA) batteries and valve regulated lead-acid (VRLA) batteries, have
been in service in electric power applications for two decades according to Butler
(2002). Nickel-cadmium (NiCd) batteries have also reached an important
maturity degree. Advanced battery technologies such as sodium-sulphur (NaS)
and lithium-ion are quickly becoming commercially available. Lithium-polymer
(Li-polymer) and nickel-metal hydride (NiMH), which have been developed
mainly for automotive use, and metal-air, are also candidate storage media.
5
Electricity Storage and Wind Energy Intermittency
Final Report
•
Flow Batteries (FB), also known as Regenerative Fuel Cells or Redox Flow
Systems are a new class of battery that has made substantial progress technically
and commercially in the last years. Flow Batteries Energy Storage (FBES)
systems have features that make them especially attractive for utility-scale
applications. The operational principle differs from classical batteries. The latter
store energy both in the electrolyte and the electrodes, so to speak. Flow batteries,
however, store and release energy using a reversible reaction between two
electrolyte solutions separated by an ion permeable membrane. Both electrolytes
are stored separately in bulk storage tanks, whose size defines the energy capacity
of the storage system. The power rating is determined by the cell stack. Therefore
the power and energy rating are decoupled, which gives the system designer an
extra degree of freedom when designing the system. Many different electrolyte
couples have been proposed for use in flow batteries. Current developments are
based on vanadium redox, sodium polysulphide / sodium bromide and zinc /
bromine.
•
Hydrogen is an immature technology but envisaged as a promising means of
electrochemical storage attracting huge interest and research funding in Europe
and the USA particularly. In a Hydrogen Energy Storage (HES) system, the
charge takes place when the electrical energy is used in an electrolyser to split
water into hydrogen and oxygen. The oxygen is usually vented to the atmosphere.
And the hydrogen can be stored in different ways. The discharge, providing the
energy release, can take place in a fuel cell or in an internal combustion engine.
One significant advantage of hydrogen as a storage option is that the energy
storage capacity input power rating and output power rating are completely
decoupled. Most aspects in the hydrogen-related technology, including generation,
storage and utilisation in fuel cells, need further development. The most severe
problem that burdens HES is the low round-trip efficiency. There are losses in the
electrolyser, storage and fuel cell. Technological breakthroughs will improve the
efficiency, but it will still remain considerably behind other competing
technologies. Despite the concerns that hydrogen arouses, hydrogen does not pose
more safety problems than other fuels. Being the lightest gas, hydrogen quickly
disperses into the environment in the event of leakage, making it less of a fire
hazard than gasoline.
Economic viability
The viability of energy storage was examined using a financial model, which
considered the variable inputs for cost and prices. The key cost inputs are the capital
and operational costs associated with constructing a storage system. The costs
associated with construction of wind farms were not included in this analysis.
Similiarly, the key benefit considered was the improvement of price in the market. In
the absence of detailed information on the market support mechanism post 2005,
proportions of green benefits were also included considering the combined price of a
traded UK Renewable Obligation type certificate and carbon taxes. Other key inputs
include the efficiency for each stage of the storage/discharge process, operating hours
per day (of charge and discharge) and cost of capital.
The model output gives the likely revenue enhancement and net present value for
systems within a range of sizes, and configurations. Figure 1 below gives the output
6
Final Report
Electricity Storage and Wind Energy Intermittency
from the economic model as a viability surface, where NPV of a hydrogen wind
system is given in terms of variance in system capital cost and system efficiency.
These two variables are the most easily influence by a concerted drive to develop this
technology. Typically with capital cost at or around €1,200/kW and system efficiency
at approximately 40%, the system becomes profitable in current market pricing
conditions.
Economic Model of Hydrogen Wind System
600
800
1,000
1,200
1,500
System Capital Cost €/kwh
1,700
0.75
400
0.55
6.0
4.0
2.0
0.0
2.0
4.0
6.0
8.0
10.0
0.35
NPV € Millions
System
Efficiency
Notes: 1. Key static inputs: 5 MW Electrolyser, with ICE Turbine Gensets, or Fuel Cells, Advanced storage medium, or full buffer
2. Input cost of wind power €0.0/MW (marginal cost), output price €40/Mwh (Nogales)
3. ICE engine uses electrolyser output at ambient pressure, output diverted for 10 hours per day (40MW/day)
4. Equipment cost from (Pritchard, Liu and proc. EERE)
5. Financing cost 7% pa., lifetime 20 years
Figure 1 Economic model of wind hydrogen systems
The results for pumped hydro are shown in figure 2. Again the key inputs to the
model are capital cost, revenue enhancement potential and system efficiency. Lower
capital costs for pumped hydro seem to be attainable, if a lower rating system is
designed for 10-50 Megawatts, rather than the 200+MW systems which require
construction of receiving reservoirs. Similarly advanced pumping and generation
technologies allow for high efficiencies in both smaller and larger plants, (greater than
60%).
7
Final Report
Electricity Storage and Wind Energy Intermittency
Economic Model of Small Scale Pumped Hydro
NPV € Millions
8.0
6.0
4.0
2.0
0.0
2.0
4.0
6.0
8.0
10.0
400
600
800
1,000
System Capital cost €/kwh
1,200
1,500
1,700
1,900
0.35
0.55
0.75
System
Efficiency
Notes: 1. Size data : 11 MW rated facility
2. Assumed output at 10 hours per day
3. Attained price enhancement of €30 per megawatt
4. Financing cost 7% pa., lifetime 20 years
Figure 2 Economic model of wind – pumped hydro
Analysis of the results from the model gives the following insights:
•
Total system capital cost is the most important variable driver of the attractiveness
of storage systems. Pumped hydro systems can be viable in certain current market
environments. This refers specifically to small scale lower cost plants with
favourable topographical conditions. However cost of storage capacity means that
they are unlikely to fully compensate on a technical basis for the intermittency of
wind. This would typically require storage capacity for up to three days of rated
output
•
Wind hydrogen systems are not yet economically viable, but expected
improvements in capital cost and system efficiency will likely change this result
over the 5-10 year term, in the context of current technology. This is based on
electrolysis and gas engine technology and excludes fuel cells.
•
Specifically hydrogen wind systems operating at above 40% system efficiency
(product of charge and discharge efficiency), can be economically viable if the
combined capital cost of the storage project is below €1,200 /kW. This applies for
currently expected market and operating conditions in markets such as the UK,
where the determinant variables are relatively well known.
•
In the Irish context, where the variables are not yet defined, the viability of storage
will be defined by the electricity price variations within any given day, and the
resulting opportunities for timeshifted price arbitrage, the charges paid by wind
8
Electricity Storage and Wind Energy Intermittency
Final Report
energy generators for reserve, the converse price that will be available for the
provision of reserve, and the value to be placed on the renewable benefit.
•
A complete model of flow battery economics and viability will await detailed
operational data. The Regenesys plant in the UK was expected to provide this in
the short term but this project has been abandoned since December 2003.
Strategy and recommendations
The short to medium term strategy focuses on the utilisation of mature electricity
storage technologies where performance characteristics and costs are known and
understood. The longer term strategy concentrates on technologies that are not yet
mature but are potentially more promising in terms of their suitability in addressing
wind energy intermittency.
The strategies focus on the storage technologies themselves and how they will operate
within the context of electricity network and electricity market developments.
Short to medium term strategy
The key elements of the short term strategy are :1. Pumped hydro resource study. A significant theoretical resource (up to
1,000MW) has been identified within the context of this study. A study is
required to determine the practicable pumped storage potential, taking into
account technical and non-technical constraints.
2. Compressed air energy storage. The potential for compressed air energy
storage should be undertaken providing details of optimum locations close to
gas generators with underground reservoirs. This will entail geological
surveying and electromechanical modifications to existing or proposed gas
fired generators.
3. System modelling. The use of storage needs to be considered in the context of
an integrated approach to dealing with wind energy interactions with the
electricity network, including specific focus on the technical issues causing
ESB National Grid to seek a moratorium on new connection agreements. This
will require the development of real time energy systems models linked to
pending grid modelling studies and incorporating the use of wind energy
forecasting. It should also consider the use of methods for addressing
intermittency other than storage (for example open cycle gas or and East West
interconnector), that fell outside of the scope of this study;
4. Grid upgrading programme. The current extensive grid upgrading programme
currently underway should be reviewed to take account of the prolific increase
and concentration in anticipated future wind energy production.
5. Demonstration Projects. The purpose of these projects is to link mature
storage technologies with wind energy to demonstrate the technical and
economic viability of the complete system. This will drive the learning curve,
reduce capital costs and increase future operational efficiencies:a. Wind + Small scale pumped hydro
9
Electricity Storage and Wind Energy Intermittency
Final Report
b. Wind + Compressed air
Long term strategy
The key elements of the long term strategy are
1. Linking wind energy storage and the hydrogen economy. A study will be
required to detail the synergies between hydrogen production in the context of
wind energy storage and the development of the hydrogen economy. In
particular the anticipated advances in hydrogen fuel cell technologies will
increase the value of hydrogen and as a result improve the economics of wind
hydrogen systems. This study will only be meaningful when data becomes
available from various detailed studies that are currently underway.
2. Demonstration projects.
a. Wind + flow battery
b. Wind + hydrogen engine
c. Wind + hydrogen + fuel cell
In summary, the energy storage sector is central to the full integration of wind energy
generation. There are a number of appropriate technologies, the most attractive of
which from a flexibility viewpoint is wind hydrogen, because of the ability to
decouple the input power, output power and storage capacity. Furthermore, wind
hydrogen systems are attractive from the standpoint of achieving zero emissions
energy. It has not yet matured from an economic perspective however and the overall
energy efficiency remains poor.
Pumped hydro systems and compressed air systems have the advantage of technical
maturity, economic viability and operational experience and are therefore viewed as a
realistic and appropriate first stage in the development of an energy storage solution
to wind energy intermittency.
10
Electricity Storage and Wind Energy Intermittency
Final Report
Table of contents
Executive summary...................................................................... 1
Context
.............................................................................................................. 1
Electricity storage ...................................................................................................... 2
Wind intermittency and storage................................................................................ 4
Economic viability ...................................................................................................... 6
Strategy and recommendations................................................................................ 9
Short to medium term strategy .............................................................................. 9
Long term strategy............................................................................................... 10
Chapter 1 The integration of wind energy in Ireland.............. 15
1.1
1.2
1.3
Projections of wind deployment in Ireland................................................. 15
Accommodating increasing wind capacity in Ireland. .............................. 19
The power sector framework: ...................................................................... 21
Trends in the power generation market ............................................................... 21
Liberalisation in the Irish electricity market .......................................................... 23
Grid upgrading plans: .......................................................................................... 25
1.4
Strategies addressing high wind penetration problems........................... 28
Wind forecasting.................................................................................................. 28
Demand Side Management ................................................................................. 29
Changes in the generation plant mix and operation ............................................ 30
Electrical energy storage ..................................................................................... 30
Chapter 2 Electrical energy storage ........................................ 33
2.1
General issues............................................................................................... 33
Benefits of storage............................................................................................... 34
Barriers to the deployment of electrical energy storage....................................... 34
Location of storage systems................................................................................ 35
2.2
Applications................................................................................................... 36
Load management............................................................................................... 36
Spinning reserve.................................................................................................. 37
Transmission and distribution stabilisation and voltage regulation ...................... 37
Transmission upgrade deferral ............................................................................ 38
Distributed generation ......................................................................................... 38
Renewable energy applications........................................................................... 38
End-use applications ........................................................................................... 39
Miscellany............................................................................................................ 41
2.3
2.4
2.5
Technical requirements of storage applications ....................................... 42
Market potential of storage applications .................................................... 43
Renewable energy storage applications .................................................... 43
Energy trading ..................................................................................................... 44
Network services ................................................................................................. 44
Scale and location of the storage systems .......................................................... 45
2.6
Research and investment in electrical energy storage............................. 46
11
Electricity Storage and Wind Energy Intermittency
Final Report
European Union research.................................................................................... 46
International Energy Agency................................................................................ 47
United States ....................................................................................................... 48
Other useful links................................................................................................. 48
Chapter 3 Storage technologies .............................................. 49
3.1
3.2
3.3
3.4
3.5
3.6
Pumped hydro ............................................................................................... 51
Compressed air ............................................................................................. 52
Flywheels ....................................................................................................... 54
Super-capacitors ........................................................................................... 56
Superconducting magnets........................................................................... 57
Batteries ......................................................................................................... 59
Applications ......................................................................................................... 59
Classical vs. advanced batteries ......................................................................... 60
Lead-acid batteries .............................................................................................. 61
Nickel-cadmium batteries .................................................................................... 63
Lithium-ion batteries ............................................................................................ 64
Sodium-sulphur batteries..................................................................................... 64
Metal-air .............................................................................................................. 65
Comparison table ................................................................................................ 66
3.7
Flow batteries ................................................................................................ 66
Vanadium Redox Flow Battery ............................................................................ 67
Polysulphide bromide flow battery ....................................................................... 68
Zinc bromine flow battery..................................................................................... 71
3.8
Hydrogen energy storage............................................................................. 72
Components of a HES......................................................................................... 73
Electrolyser.............................................................................................. 73
Power generation: fuel cells................................................................................. 75
Reversible fuel cells............................................................................................. 77
Storage and compression.................................................................................... 77
Power electronics ................................................................................................ 79
Synergy: The hydrogen economy........................................................................ 79
Projects ............................................................................................................... 80
3.9 Power conditioning subsystem and balance-of-plant............................... 81
3.10 Costs ............................................................................................................ 82
3.11 Comparison of electricity storage technologies........................................ 85
Chapter 4 Storage to accommodate wind energy in Ireland . 89
4.1
4.2
4.3
Selection of technologies for different storage applications ................... 89
Experience in the use of storage for the integration of RE ...................... 91
Assessment of different technologies........................................................ 93
Pumped hydro ..................................................................................................... 94
Compressed air ................................................................................................... 94
Flywheels............................................................................................................. 95
Supercapacitors and supermagnets .................................................................... 95
Batteries .............................................................................................................. 96
12
Electricity Storage and Wind Energy Intermittency
Final Report
Flow batteries ...................................................................................................... 96
Commercialisation characteristics ........................................................... 97
Economic comparison ............................................................................. 97
Activity in the RE sector......................................................................... 100
Hydrogen........................................................................................................... 101
4.4
Conclusions................................................................................................. 101
Chapter 5 Economic viability of storage options ................. 103
5.1
Introduction and summary......................................................................... 103
Summary economic findings:............................................................................. 104
5.2
Costs: Present and Future ......................................................................... 106
Economic Underpinings..................................................................................... 106
Potential for renewable energy dispatch............................................................ 107
Categorisation of Costs and Prices ................................................................... 108
Current Capital costs ......................................................................................... 109
Operating costs and costs per discharge; [4]..................................................... 110
Input and output energy costs: [Platts (various) and CER (2003d)]................... 111
Cost improvements areas to examine [Baker (2001), Altmann, Niebauer et al
(2000), Nicoletti (1995)] ..................................................................................... 111
5.3
Financial model ouputs:............................................................................. 113
Benefits in the context of a liberalised electricity market.................................... 113
Wind hydrogen system ...................................................................................... 114
Wind pumped hydro system .............................................................................. 116
External benefits................................................................................................ 117
Chapter 6 Strategy................................................................... 119
Short to medium term strategy .......................................................................... 119
Long term strategy............................................................................................. 120
References
............................................................................ 123
Appendix 1. Wind farms with planning permission .............. 131
Appendix 2. Variable used in financial model ....................... 135
Variable ............................................................................................................. 135
Description......................................................................................................... 135
Typical Range.................................................................................................... 135
13
Electricity Storage and Wind Energy Intermittency
Final Report
14
Final Report
Electricity Storage and Wind Energy Intermittency
Chapter 1 The integration of wind energy in Ireland.
1.1
Projections of wind deployment in Ireland.
Until recently, the rate of wind energy deployment in Ireland has been relatively
modest at approximately 20MW installed capacity on average per year since 1997.
This rate increased more recently and the total power connected by May 2004 from
wind farms was 190MW, with a further 44MW nearing final connection. The location
of these wind farms is shown in figure 1.1.
However this recent acceleration rate is likely to continue as evidenced by recent data
(CER 2004) on the number of signed agreements, live offers and applications for grid
connections, summarised in Table 1.1
Connected
Signed
Agreements
Live offers
Applications in
process
Applications
being checked
Transmission
(MW)
39
Distribution
(MW)
151
Total
(MW)
190
Cumulative
Total (MW)
190
379
254
633
823
0
10
10
833
207
532
739
1572
407
220
627
2199
Table 1.1 Status of wind farm grid connection agreements May 04
This summarises the situation regarding grid connections and could be regarded as the
best indicator of the future installed capacity for the next few years. Adding to the
190MW currently connected the amount of wind power not yet installed but with
signed connection agreements raises the total to 823MW, with a further 749MW
within the process and an additional 627MW for which applications were being
checked in May 2004. This represents a cumulative total of 2199MW, although it is
unclear how many of these wind farms have secured planning permission and the
necessary finance to enable construction. A survey carried out in 2004 (CER 2004)
concluded that 661MW would be likely to connect to the system by mid 2006,
bringing the total installed wind capacity to 851MW. This represents 14% of
projected total generation capacity in 2006 (ESBNG 2003).
It should be pointed out that despite the recent dramatic increase in activity, it is
almost certain that the target in the Green Paper on Sustainable Energy will not be
met. The target set was for an additional 500MW of renewable generated electricity to
be delivered by 2005, i.e. before January 1 2005. Excluding AER III wind farms that
fell outside this target, 92MW has been delivered to date and based on the timeframe
envisaged for connections (ESB National Grid 2003) an additional 308MW will be
connected by the end of 2004, delivering thus 400MW of the 500MW total. Based
again on accepted connection offers, it is envisaged that an additional 186MW of
wind energy capacity will be installed during 2005.
15
Electricity Storage and Wind Energy Intermittency
Final Report
Figure 1.1 Map of wind farm sites in Ireland
Table 1.1 above does not give the complete picture of potential wind energy
deployment however, as there may be wind farms that have secured a grid connection
agreement but do not have planning permission, and vice versa. Table 1.2 gives the
16
Electricity Storage and Wind Energy Intermittency
Final Report
total capacity with planning permission (in November 2003) on a county-by-county
basis.
County
Wind Power Currently
Installed / MW
Additional Capacity with Portion of which under
PP / MW
appeal MW
Roscommon
27.2
10.56
3
65.07
23.34
3.975
11.9
0
0
0
0
9.4
25
0
0
0
0
10.94
326.4
247.245
227.5
206
198.7
139.81
121.3
77.75
74.16
54.6
52.5
48.36
14.4
12
5
5
2.55
0
323.4
6
58
0
4
0
0
0
0
0
52.5
0
3
0
0
0
0
0
TOTAL:
190.335
1813.275
446.9
Mayo
Cork
Cavan
Donegal
Kerry
Galway
Wexford
Limerick
Tipperary
Clare
Carlow
Leitrim
Sligo
Meath
Kilkenny
Offaly
Wicklow
Table 1.2 Installed capacities of wind farms with planning permission by county
It is clear from table 1.2 that, based on the number of wind farms with planning
permission, there is the potential for a cumulative installed capacity of at least
2004MW of on-shore wind, which compares well with the 2199MW total figure
based on connection agreement applications (that includes on-shore and off-shore
wind farms). In addition to this 2004MW, there is significant interest in developing
offshore wind energy, potentially accounting for an additional 2,000MW, according
to Garrad Hassan, ESBI & UCC (2003).
This highlights the need for the development of a co-ordinated wind energy strategy,
which will achieve a balance between the technical constraints relating to grid
integration and the contribution that can be made by wind energy to meeting national
objectives arising from commitments under the Kyoto Protocol, EU Directive
2001/77/EC on the promotion of electricity from renewable energy and the drive for
sustainability in the energy sector. The current consultation process on policy, targets
and measures for Ireland (DCMNR 2003) will need to feed into a strategy to address
the mismatch between projects with planning permission, grid connection agreements
and market access as all three are necessary for a wind farm to be built. The locations
of potential wind farm sites with planning permission are illustrated in figure 1.2.
The increase in the concentration of wind farms in certain locations is apparent from a
comparison of figures 1.1 and 1.2 and it is important to note that wind farm
concentration exacerbates the impacts of wind intermittency (RESG 2000). It should
also inform grid upgrading programmes to facilitate increased wind penetration.
17
Electricity Storage and Wind Energy Intermittency
Final Report
Figure 1.2 Map of wind farm sites in Ireland with planning permission
Table A.1 in Appendix 1 provides a list of all wind farms with planning permission
and figure A.1 shows the locations of existing operational wind farms.
18
Electricity Storage and Wind Energy Intermittency
1.2
Final Report
Accommodating increasing wind capacity in Ireland.
Garrad Hassan, ESBI & UCC (2003) reported on the impacts of increased wind
penetration on the Irish electricity networks, identifying two fundamental types of
limiting factors for the connection of wind to the network as follows;
Type 1 ……. Transmission planning criteria
Under existing transmission planning criteria all generation must be considered as
firm i.e. it must be able to continue to operate in the event of any one of a defined set
of contingencies on the transmission system. This affects wind farms as they may not
connect until sufficient grid reinforcement is in place to allow for firm connection of
the wind farm. The authors suggested an alternative method of dealing with
contingencies that would allow wind farms to connect on a non-firm basis and thus
avoid delays associated with deep grid reinforcement. The threshold at which this was
deemed necessary was well under 1000MW.
Given that wind farms with a combined installed capacity 823MW have received
connection agreements on a ‘firm’ basis without the need for significant deep
reinforcement indicates that this constraint was not as significant a barrier as
suggested by the authors.
Type 2 …….. Curtailment at times of low load.
A key assumption underpinning the study was that as the output of wind generation
increases, the output from existing and planned fossil fuel power stations is reduced,
but not shut down. There is a limit to the part-load operation of these stations for
technical and efficiency reasons. Once this limit is reached, the study required that the
output from wind generation be curtailed.
Additional problems associated with transient and voltage stability were also
identified in the study. These have not been properly analysed at present and
considerable additional dynamic grid modelling studies are required. Due to absence
of understanding relating to these key issues, ESB National Grid (2003) made a
request to the Commission for Energy Regulation (CER) to suspend grid further
connections until the technical issues regarding security and stability of the power
system are fully resolved.
CER agreed on an exceptional basis to a moratorium on connection offers until the
end of 2003 and invited the wind industry to submit comments on these emergency
measures. CER also proposed that ESB-National Grid host a forum to discuss the
issues before the CER takes a final decision on future grid connections. The forum
took place in Citywest Hotel Dublin on 17th December 2003.1
1
Presentations from this forum are available from
http://www.eirgrid.com/EirGridPortal/DesktopDefault.aspx?tabid=Wind%20Forum%20Present
ations%2017th%20December
19
Electricity Storage and Wind Energy Intermittency
Final Report
Following the forum, ESB National Grid requested an extension of the moratorium
until March 31st 2004. CER granted this extension, on the basis that a number of
specific issues be undertaken within the 3 month extension period, namely that
•
the programme addressing the requirements of the Grid Code for Wind be
accelerated and the need for interim reporting and consultation be addressed;
•
interacting issues with the Distribution Code arising from the Grid Code for Wind
review be resolved;
•
a survey be conducted of current connection offers to the transmission and
distribution systems to better assess their projected timeframes and potential
impact on the system;
•
issues regarding the constraining of wind farms be examined;
•
the differences in the connection offer processes between the transmission and
distribution systems be clarified and reconciled;
•
a detailed programme and timetable for the modelling of wind generation plant
and the impact of greater penetration of wind on the transmission system be
produced;
•
a working group containing the Commission, Transmission System Operator
(TSO), Distribution System Operator (DSO) Sustainable Energy Ireland (SEI) and
the Irish Wind Energy Association (IWEA) be formed to monitor progress on these
and related issues;
•
a workplan detailing the programme of work for the next three months be
submitted to the Commission and published in January 2004.
There are clear overlaps between these conditions and the recommendations made by
Garrad Hassan, ESBI & UCC (2003) that;
•
The TSO should define more closely their concerns about dynamic issues.
•
Further work should be carried out to establish with the wind turbine
manufacturers what their products can do and are expected to do in the near
future. Effectively this requires that WTG’s seeking connection in Ireland should
be able to demonstrate compliance with Grid Codes.
•
An assessment of the risk associated with the expansion of wind generation on the
system is required.
In brief, Grid Codes need to be developed, the interaction of the WTG’s with the grid
need to be assessed and the overall dynamic performance of the system with high
levels of wind penetration requires modelling.
The development of the Grid Code for Wind Energy was accelerated and consultation
is currently taking place n the draft code (ESBNG 2004). Dynamic models have
become available for certain wind turbines and modelling is progressing, The CER
has proposed lifting the moratorium subject to a wind farms connecting to the system
complying fully with the finalised code (including fault ride through provisions) and
20
Electricity Storage and Wind Energy Intermittency
Final Report
providing turbine models with the grid connection application to allow modelling of
system impacts to be carried out.
1.3
The power sector framework:
Trends in the power generation market
There have been a number of significant trends in electricity generation in Ireland
over the past decade, including :•
•
•
•
•
Significant growth in electricity generation
Increased penetration of natural gas
Improved energy and CO2 efficiency of generation
Shortages of supply
Growing influence of environmental constraints
The most significant trend has been the significant levels of growth in electricity
consumption, prompting by the dramatic increase in economic growth (measured by
Gross Domestic Product, GDP). According to Howley, M. & Ó Gallachóir, B. (2002)
the 7.2% average annual growth in GDP during the 1990s was the primary driver for
energy consumption growth. During the period 1990 – 2001, the average annual
growth in final demand for electricity was 5.3%, demonstrating that electricity
demand is not directly coupled to economic growth. The annual growth in electricity
generation in the same period was 5.1%, indicating some improvement in reducing
line losses.
Much of this increase in generation has been met by increased use of natural gas, oil
and wind energy has also made a contribution. Natural gas use for electricity
generation has grown by 120% between 1990 and 2001. It accounted for 35% of the
electricity generation fuel mix in 2001, compared with 27% in 1990, making it the
most significant fuel in electricity generation. Oil has also grown significantly, from
11% of the fuel mix in 1990 to 22% in 2001. Wind energy still contributes a small
contribution to our overall electricity requirements but this is set to change
dramatically in the short term, as was shown in section 1.1.
These changes in the fuel mix, in addition to technological changes (the trend towards
more efficient combined cycle gas generation), have increased the efficiency of the
electricity system. Defined as the final consumption of electricity divided by the fuel
inputs required to generate this electricity, efficiency has increased from 33% in 1990
to 35% in 2001.
The growth in electricity consumption prompted a need for new capacity, in particular
following the significant increase in the latter part of the decade (average annual
growth between 1998 and 2000 was 7%). ESB National Grid (2002) reported that a
460MW combined cycle gas plant at Poolbeg had its first full year of operation in
2000 followed by the 118MW Edenderry peat plant that was commissioned in 2000.
In order to address generation adequacy, ESB leased five 22MW emergency
generators for winter 2000. In winter 2001, ESB procured an additional two 22MW
distillate-fired emergency generators.
21
Electricity Storage and Wind Energy Intermittency
Final Report
During 2002 two additional combined cycle gas plants were commissioned, Dublin
Bay Power (408MW) and Huntstown (343MW), significantly adding to our
generating capacity. Over the last few years, there has also been a number of peat
fired generating plants decommissioned, at Ferbane (90MW), Rhode Island (80MW),
Caherciveen (5MW) and part of the Lanesboro plant (45MW).
Looking ahead, the remaining pre 2000 plants at Shannonbridge (125MW),
Lanesboro (the remaining 40MW) and Bellacorrick (40MW) are also all due to close
by the end of 2004. Two new peat plants are planned to be commissioned during 2004
to replace the capacity lost, Lough Ree Power at Lanesboro (100MW) and West
Offaly Power at Shannonbridge(150MW).
In addition, a number of measures have been introduced to offset anticipated
shortfalls in generation including
•
•
•
•
introduction of a Winter Peak Demand Reduction Scheme (WPDRS) from
November to February – providing a winter demand reduction incentive within
the new market structure, designed to reduce peak load at the time of highest
demand in the year;
the purchase 170MW capacity from the Ballylumford generation station in Co.
Antrim from 2003 to 2006 through a contract between ESB and NIE;
the CER Capacity 2005 competition, designed to facilitate the entry of up to
531MW of new, independent generation by the end of 2005. The successful
bidders will enter into power purchase agreements for a maximum of 10 years
with ESB.
the installation in December 2003 of two 52MW open cycle gas fired plants in
Aghada, Co. Cork and Tawnnamore, Co. Mayo. These are peaking capacity
plants, to be available until the new plant built under the CER Capacity 2005
competition is delivering electricity.
The impact of environmental constraints on electricity generation is growing
significantly. As DELG (2000) point out in National Climate Change Strategy, a
target of 5.7 Mt CO2 per annum reduction below business as usual projections is set
for the electricity supply industry. This represents more than one third of the total
emission reductions (15.5 Mt). The bulk of this is to be achieved by closing the
Moneypoint coal fired station and replacing it with a combined cycle gas plant (3.4
Mt) and increased penetraion of renewable energy (1 Mt).
Since this has been published the EU Emissions Trading Directive 2003/87/EC has
been agreed (October 2003) that will establish a cap and trade system for greenhouse
gas emissions including most thermal power plants in Ireland.
The National Emissions Ceiling Directive 2001/81/EC sets upper limits (by 2010) for
the four pollutants responsible for acidification, eutrophication and ground-level
ozone pollution (SO2, NOx, VOCs and ammonia), but leaves it largely to the Member
States to decide which measures to take in order to comply. Based on the provisions
of the Directive, Member States are obliged to report each year their national emission
inventories and projections for 2010 to the European Commission and the European
22
Electricity Storage and Wind Energy Intermittency
Final Report
Environment Agency. They shall also draw up national programs in order to
demonstrate how they are going to meet the national emission ceilings by 2010.
The Large Combustion Plant Directive 2001/80/EC focuses specifically on limiting
the emissions of SO2 and NOx from large combustion plants, including power plants.
In order to meet the provisions of these latter two Directives, ESB (2003) proposes to
install Selective Catalytic Reduction (SCR) to reduce NOx emissions and flue gas
desulphurisation technology to reduce SO2 emissions from the Moneypoint coal fired
plant. In addition, ESB proposes to reduced use of oil and the introduction of low
sulphur oil and coal into the fuel mix.
Liberalisation in the Irish electricity market
A further key policy development that impacts significantly on the power sector
framework has been the introduction of electricity market liberalization, in February
2000 allowing the direct sale of electricity to customers. Liberalization in the Irish
electricity market is enshrined in Electricity Regulation Act, 1999, which sets out to
implement EU Directive 96/92/EC concerning common rules for the internal market
in electricity. This allows independent electricity generators and/or suppliers to
contract directly with designated customers for the supply of electricity.
The Directive required that approx. 28% of the market be opened up to competition in
2000, increasing to 33% by 2003 with a review of further opening in 2006. In fact,
Ireland has gone further than this with approximately 30% of the market opening
initially, rising to 40% in 2002 and set to increase to 56% in 2004.
Directive 96/92/EC was updated in 2003 with Directive 2003/54/EC concerning
common rules for the internal market in electricity. This provided for full opening to
the industrial and commercial customers by July 2004 and full market opening to all
customers by July 2007. The Irish market is due to be fully open by July 2005, well in
advance of the EU deadline, according to CER (2003)
The initial and current interim market is based on bilateral contracts between
electricity suppliers and customers. Under the Act, large electricity consumers above
a certain threshold of annual consumption can choose their electricity supplier. In
addition those who supply electricity from renewable energy sources can sell directly
to all final customers. As a result, brown electricity suppliers can sell only to large
customers but green electricity suppliers can sell to customers of any size. Green
electricity suppliers thus have had access to the sections of the market which pay most
for electricity (commercial and domestic customers). This has clearly good news for
wind energy, but the challenges facing the sector should not be underestimated.
The green electricity supplier will need to source green electricity to meet the demand
of the customer base (which needs to be established). Already new players are
entering the market to act as brokers, with the aim of buying from a portfolio of green
electricity generators and selling to a portfolio of customers. Airtricity Ltd., e Power
Ltd., ESB Independent Energy Ltd. and E.Co – The Electricity Company Ltd. have
secured licenses to supply green electricity.
23
Electricity Storage and Wind Energy Intermittency
Final Report
The details of market trading will change considerably with the introduction of new
market arrangements from 2005. Both the current and new arrangements pose
particular challenge for intermittent renewable energy sources such as wind energy,
due to the half hourly trading period that is in place.
Currently the scheduling, trading and settlement are based on half-hourly intervals.
Imbalances in electricity scheduled compared with that traded, are dealt with through
a balancing market. In the case of a wind farm, for example, at certain times, more
electricity is produced and dispatched than actually consumed by the green electricity
customers it is destined for. The excess electricity in this half hour may then be sold
to another generator who had a shortfall in the same period provided this is done
within 7 days of the trading day. Otherwise ESB (Generation) buy this amount of
electricity at the ‘spill’ price (ESB’s avoidable fuel price up to an initial tranche and
thereafter the avoidable fuel cost of the best new entrant).
Equally, less electricity can be produced and dispatched from the windfarm, than was
actually consumed by the customer(s). The shortfall in electricity in this half hour
must then be purchased from another generator who had an excess in the same period
provided this is done within 7 days of the trading day. Otherwise ESB (Generation)
will sell this amount of electricity at the ‘topup’ price (which should average out over
the year to the estimated full cost of a best new entrant).
Electricity from intermittent renewable energy sources relies more on the balancing
market than, for example, gas generated electricity, which being more predictable,
will incur less imbalances and not be as dependent on the balancing market. This
affects the economics of trading green electricity and requires suppliers to incorporate
this into the sale price sought for the direct sale of wind generated electricity. This is
offset currently by the favorable conditions that exist for wind energy suppliers,
namely access to the full market and access to unlimited top-up in a given half hour.
A further key challenge that arises for green electricity generators selling to green
electricity suppliers compared to those with an AER contract is the absence of a fixed
price 15 year power purchase agreement which offers significant comfort to financiers
in the “AER market”. The supplier must seek customers who are willing to agree to
purchase green electricity at predetermined rates for a certain time period. It seems
likely that a 3 year contract would be the maximum that a green customer would sign
up for. This will clearly have implications for the financial risks and the availability
and cost of finance in the absence of a guaranteed sales mechanism.
The market mechanism that will replace this interim bilateral trading market will be a
centralized wholesale electricity market. This new system will be a mandatory
centralised pool (“the spot market”) requiring all electricity exported to or imported
from the transmission system or distribution system to be sold to and bought from the
SMO (System Market Operator, within ESB National Grid).
In this centralised market, all power that is generated will receive the Market Clearing
Price, which may be different in different locations. Generators signal with their
offers when and how much they would like to generate. The market is cleared based
on these offers and dispatch instructions issued accordingly. The market clearing price
24
Electricity Storage and Wind Energy Intermittency
Final Report
is set by the highest offers accepted by the market. In a LMP (Locational Marginal
Pricing) market the market clearing price is set for each node of the network.
For renewable plant reliant on intermittent power sources such as wind, offering into
the market and adhering to dispatch instructions carries particular problems. The
problems they face are both technical and commercial.
In order to clarify the situation with regard to renewables in the new market, CER
(2003) propose that
1. wind turbines below 10MW and wind farms below 30MW will be able to
register with the SMO as “not-dispatchable” and so may not be subject to
dispatch instructions.
2. there be no market floor price for renewables or CHP, other then the general
market price floor of negative VoLL (Value of Loss of Load). If it is required,
renewables and CHP should be compensated outside of the market
arrangements, through an additional support mechanism.
3. all generators be liable for the cost of reserves in line with a ‘causer-pays’
principle. These costs will be allocated in proportion to the requirements for
reserves that are deemed to be due to each generating unit.
The new market poses significant and different challenges for wind energy than the
current market. The use of financial tools (contracts for differences and other hedging
instruments) will be very important. In addition, unresolved issues relating to
Financial Transmission Rights, the detail on the operation of the reserve market and
how priority dispatch for renewables will all have a key bearing on the viability of
wind energy in the future.
Grid upgrading plans:
As noted by the International Energy Agency (2003), the transmission system
comprises over 5,800 km of high voltage lines operating at 110kV, 220 kV and 400
kV. The national grid was initially established as a 110 kV network but, as the
demand for electricity grew, the 220 kV and 400 kV networks were added. The
transmission system also includes over 100 high voltage transformer stations where
voltage is reduced fur use in the local distribution lines at voltages of 38 kV, 20 kV
and 10 kV. The distribution network includes about 80,000 km of overhead wires
and underground cables.
As electricity demand has grown in the past decade, the transmissions system has
become strained, prompting the urgent need for system upgrade. A major
refurbishment and expansion programme running from 2001 to 2005 is underway.
This programme increases annual capital expenditure on the transmissions and
distribution networks by a factor of three, contributing to the recent electricity price
increases. Over €2.6 billion is being invested in the high voltage and low voltage
networks, particularly in the counties along the southern and western coasts. Over €
820 million will be spent on transmission, over € 1 billion on distribution renewal
(including conversion of 50% the 10 kV network to 20 kV), and over € 665 million on
distribution reinforcement.
25
Electricity Storage and Wind Energy Intermittency
Final Report
This upgrading of the electricity networks will improve the system and thereby
facilitate the accommodation of wind energy. In addition, separate measures have
been identified by the Renewable Energy Strategy Group (2000) to facilitate wind
energy.
The Group recommended a short and medium term approach. In the short term, the
absence of sufficient capacity and the financing arrangement for additional capacity
were to be addressed. Specifically, the Group recommended that
1. some funds identified in the National Development Plan should be released to
finance the additional costs of delivering additional capacity at designated
locations, which the Department of Public Enterprise will supervise. The
extent of upgrading will depend on perceived demand and subsequent
connection charges will be proportional to the capacity connected. For
example, if the perceived demand is 50MW in a particular area, a 5MW wind
farm would be charged 10% of the infrastructural investment instead of the
100% charge currently applied. As connections are made to the distribution
and transmission networks the charge is remitted. These funds should be
recycled on the same basis, so long as additional demand can be predicted
under reasonable assumptions,
2. CER takes wind energy into account when deciding on plans to upgrade the
network.
In the National Development Plan Economic and Social Infrastructure Operational
Programme, € 184.5 million is earmarked for sustainable energy initiatives.
Regarding the first recommendation, the Department of Communications, Marine and
Natural Resources (2002) established a Steering Group in 1999 to oversee the
implementation of the Grid Upgrade Development Programme for Renewable energy.
This Group reported in September 2002 suggesting a mechanism addressing the
challenge that existed for developers where they must raise the entire capital
expenditure for any upgrade forming part of a potentially shared connection with
money subsequently remitted as others connect to the facility. The Steering Group
concluded that
•
The grid upgrades should be planned by reference to perceived demand for
shared infrastructure;
•
Perceived demand should be based on clusters with two or more projects
with full planning permission intending to connect to the upgrade;
•
The prioritisation of clusters for investment support should operate on a first
come first served principle subject to compliance with minimum
requirements with a fall back selection criterion in the event of simultaneous
applications exceeding the available fund;
•
The first come first served principle should apply to any project compliant
with the qualifying criteria, at that time;
•
Project developers should be charged under reasonable assumptions for the
capacity reserved as a proportion of the grid upgrade built.
CER (2003) determined that
26
Electricity Storage and Wind Energy Intermittency
Final Report
in line with its duty under the Electricity Regulation Act (‘the Act’) to have
regard to promoting the use of renewable forms of energy, to support the
funding of the programme through TUoS charging. The Commission
acknowledges that this proposal leads to preferential access arrangements for a
number of renewable applicants. However, the Commission does not believe
that this constitutes unfair discrimination.
The scheme is not only underway, it does not require the NDP funding as it will be
integrated into TuoS charging.
The Group further recommended in the short term that
In addition, clear positive changes have been brought about, as a result of the
work carried out by the Working Group on Grid Connection Issues Relating to
Renewable Energies. Further improvements can be made, and in this regard the
Strategy Group endorses the following recommendation from the final report of the
Working Group:
•
the four studies, detailed below, which were commenced should be completed
as soon as is practicable. All test sites should be set up by mid 2000, and
evaluation of the collected data should be completed by mid 2001.
1.
the use of MV (medium voltage) voltage regulators with line compensation
to counteract excessive voltage rise;
2.
the use of high sensitivity over-voltage relays at embedded generator sites
policing voltage levels during load/generation variations;
3.
the testing of modern inverter technology in variable speed wind turbines to
assess harmonic performance and power factor control;
4.
the use of power supply monitors at generator sites recording voltage,
power flows, harmonics, flicker, etc.
The NDP funding not required for the short term strategy may now be used to deliver
a key element of the medium term strategy of the Renewable Energy Strategy Group
funding. The medium term strategy recommended a continuation of the mechanism
described above and that
In addition, where strategic wind energy sites are identified which require
additional transmission infrastructure then such grid upgrading should be fully
funded from remaining funding available under the National Development Plan.
Once built, however, this network extension will be available to all generators
in a non-discriminatory fashion in line with national policy.
There is no current publicly available information regarding progress has been made
in this element of the strategy.
27
Electricity Storage and Wind Energy Intermittency
1.4
Final Report
Strategies addressing high wind penetration problems
Wind forecasting
Wind forecasting does not in itself overcome the problems associated with wind
intermittency. However, an accurate forecasting tool can facilitate transmission
system operators to accommodate higher wind penetrations by planning for other
generation available at times when it is known the wind is not available and equally
curtailing other plant when the wind will be available.
There are many different wind energy forecasting techniques at various stages of
development throughout Europe and the U.S. The simplest form of prediction is based
on the persistence method (assuming that the current value persists) but this is
acceptable only for a one hour horizon. Forecasts based on Numerical Weather
Prediction (NWP) models can be used for much longer time frames of up to 48 hours
but the accuracy of the predictions is extremely variable and is very dependent on the
type of the prevailing weather system. Errors can range from 10 percent for 10 hours
ahead to 30 percent for 48 hours during stable meteorological conditions to 200
percent during unstable conditions. One of the most recent developments is called
ensemble forecasting which not only gives more accurate wind speed predictions but
also quantifies the level of uncertainty associated with the given forecast. This
approach recognises the fact that there are some periods when it is not possible to give
an accurate prediction and this uncertainty when quantified is an important piece of
additional information for the TSO. This system is being developed by the
Sustainable Energy Research Group (SERG) at University College Cork and has been
tested by Eltra in Denmark with very promising initial results. On the strength of
these results, SERG is currently co-ordinating an EU 5th Framework supported project
called HONEYMOON (a High resOlution Numerical wind EnergY Model On- and
Offshore forecasting). The goal of the project is to develop an ensemble forecasting
tool using of a numerical weather prediction model that is designed specifically for
use in a real-time wind energy prediction. This system is being tested in Ireland
through a collaborative SEI funded project between SERG and ESB National Grid.
Since 1999, Eon Netz in Germany has been using a forecasting tool based on artificial
neural networks which was developed by ISET, the German Institute for Solar Energy
Technology. These networks are collections of mathematical models that can be
trained to simulate the complex weather patterns using historical data.
Wind energy forecasting has both technical and economic benefits. The main
technical benefit is related to the load management of power stations which have to
adjust to accommodate the level of wind power on the grid system at any given time.
Advance information on the production from wind farms enables the TSO to plan
production from the fully dispatchable plant and will reduce inefficiencies caused by
enforced part load operation.
The economic benefits are significant in fully liberalised electricity markets where
TSO’s have to pay for reserve power. For example Eon Netz is required to schedule
wind power as virtual base load the day ahead and they do this based on a forecast
with a provision for reserve power from firm capacity to balance out any discrepancy
between projected output and actual delivery. Reducing the difference between
28
Electricity Storage and Wind Energy Intermittency
Final Report
projected and actual delivery of wind power through good forecasting reduces
purchases of balancing power.
Clearly wind energy forecasting will not solve the problems of associated with the
variability of wind. It will reduce however, the uncertainty regarding the wind power
production and facilitate more efficient power station load management. In any future
system, which aims at optimising the deployment of the wind energy resource in
Ireland, forecasting should play an important role integrated with appropriate energy
storage technology.
Demand Side Management
One important way that electricity companies can reduce their GHG emissions is
to reduce energy consumption among customers. This approach is referred to as
Demand Side Management (DSM). Instead of building new power plants to
respond to increasing customer demand, electricity producers can also try to
reduce their customers' demand for power by offering special programs for
businesses, industry, public institutions and domestic users. To determine the
success of such programs, the costs and benefits of DSM opportunities, including
any future financial value of greenhouse gases (GHG) emissions reductions,
should be directly compared with the costs and benefits of building new power
plants and transmission lines.
DSM programs aim to achieve three broad objectives:
•
Energy conservation: DSM programs can reduce the overall consumption of
electricity by reducing the need for heating, lighting, cooling, cooking energy and
other functions. For example, adding insulation to a building can help reduce the
need for heating in winter and cooling in summer.
•
Energy efficiency: DSM programs can encourage customers to use energy more
efficiently, and thus get more out of each unit of electricity produced. For
example, energy-efficient light bulbs provide the same amount of light but use
significantly less energy than conventional units.
•
Load management: DSM programs allow generation companies to better manage
the timing of their customers’ energy use, and thus help reduce the large
discrepancy between peak and off peak demand. For example, utilities can
interrupt industrial power supplies temporarily during periods of high demand,
and/or store power during periods of low demand for later use, when demand is
high. Although this approach does not reduce the overall consumption of
electricity, it can reduce the need to build new power plants simply to serve
customers during periods of peak demand. Load management can also reduce
GHG emissions associated with using fossil fuels to meet those peak electrical
demands.
Demand side management is well developed in the Irish situation, however it is
unlikely that DSM provides a full solution to the issues around high penetration of
renewable energy. Any move to institute a full DSM programme targeted improving
the dispatchability of wind, and effect load levelling, to improve inherent capacity
credit, should be specifically agreed with the TSO.
29
Electricity Storage and Wind Energy Intermittency
Final Report
Changes in the generation plant mix and operation
In the Irish context one of the limiting factors affecting the non dispatchable nature of
wind energy is the comparatively high baseload requirement of the conventional
thermal plant. Plants such as Moneypoint have a requirement to run at significant
capacity levels to be efficient and therefore cannot be easily ramped down as
renewable energy comes on stream.
In pure economic terms the move to significantly more variable electricity production
from CCGT plant is in the favour of wind penetration as CCGT can in fact be more
variable in response to improved generation capacity from renewable sources.
Again as in the consideration of Demand Side Management any substantial move to
include a revision of operating tactics and strategy for conventional generation plant
should be in the context of cooperation between the TSO, conventional and renewable
energy generators. Within this framework there are a number of market mechanisms
which can enhance the capacity credit open to wind energy (e.g. carbon costs, PSO
hand off etc.)
The diversion of traditional pumped hydro storage facilities to virtual wind energy
storage devices is one obvious way for increasing the dispatchability of wind energy.
This may become more compelling as the requirement to utilise pumped hydro for
short term applications like power quality and load levelling is diminished both by the
increased sophistication of grid operation and the increased availability of CCGT
capacity, within the market context, to provide these ancillary applications.
Electrical energy storage
The fundamental issue surrounding wind energy integration and is its intermittency.
As such no grid operator, especially not an isolated grid operator can accept unlimited
amounts of wind energy to its grid. Strategies such as Demand side management,
flexible operation of other generation plant, large interconnected grids, all provide
some leeway for wind energy penetration. In some cases such as Denmark significant
penetration of wind energy has been possible, because of concerted action by the Grid
operators and availability of interconnects and alternative sources of energy (in
Denmarks case Norwegian Hydro power is an important flexible input.)
However irrespective of load levelling and capacity sharing the fundamental
requirement to improve the reliability and predictability of wind energy requires some
solution which can store the energy when it is not required (and has low value) and
resupply the electrical energy when needed.
Absolute limits on energy storage capacity, and high cost per kWh of storage medium
installation, imply that pumped hydro (PHES) and flow battery storage (FBES), may
improve the availability of wind energy, but do not address the issue of multiday
unavailability of wind (meteorlogical calms often ‘blocking highs’ in a European
context). Notwithstanding this, both PHES and FBES industries have made
substantial progress, and the development of price arbitrage and high storage to output
rating configurations by a number of companies, mean that both technologies are
potentially technologically and financially viable solutions to wind intermittency in
30
Electricity Storage and Wind Energy Intermittency
Final Report
the short term. In the Irish context small scale pumped hydro seems particularly
promising in combination with individual or grouped windfarms.
Wind hydrogen systems propose a fundamental decoupling of energy production,
storage and regeneration. A number of early installations, and significant research
activity point to a number of opportunities:
•
Wind hydrogen in combination with significant storage can technically be viewed
as dispatchable generation capacity on a par with other sources of generation
•
Large strides have been made to address the system cost of wind hydrogen
systems and the combined cost2 and efficiency metrics of approximately
€1,500/kW and 45% respectively are points where a wind hydrogen system would
be economically viable
•
Economic viability is driven by four important factors:
1. The price enhancement available between variable charge cost
(foregone price) and discharge attained price
2. The average time period when the system can discharge to an ehanced
revenue environment
3. The capital cost of the facility (charging equipment, storage
equipment, and discharge equipment)
4. The efficiency of the system
Within this framework, Wind Hydrogen systems meet the technology specifications
to deal completely with wind intermittency, but are still not proven to be
economically viable. The interplay of market factors, price arbitrage and capacity
credits may have the effect of making fully integrated Wind Hydrogen viable at
present. However the potential for further economic benefit from incremental system
cost and efficiency improvements highlighted.
2
excluding the capital costs of the wind farm
31
Electricity Storage and Wind Energy Intermittency
Final Report
32
Electricity Storage and Wind Energy Intermittency
Final Report
Chapter 2 ELECTRICAL ENERGY STORAGE
2.1
General issues
The electrical energy storage concept has become a controversial issue in the last
years. Many questions arise in the electricity sector: Why is energy storage needed?
What are the alternatives? How much do storage systems cost and how much added
value does a storage system provide? Will storage contribute to the increased
utilisation of renewables?
The storage issue must be viewed in the frame of a changing electricity sector.
•
Restructuring of the electricity market
•
Growth in new/renewable energy sources
•
Increasing reliance on electricity and demand for higher quality power
•
Move towards distributed generation
•
More stringent environmental requirements
As part of these changes, there are growing pressures to operate the electrical network
more efficiently whilst still maintaining high standards of reliability and power
quality. The accommodation of renewable generation and ever more stringent
environmental requirements are combining strongly to further influence electricity
companies’ decisions on how they should be developing their future network designs.
With these driving forces as a backdrop, the rapidly accelerating rate of technological
development in many of the emerging electrical energy storage technologies, with
anticipated system cost reductions, now makes their practical application look
attractive.
Energy storage is not a new concept in the electricity sector. Utilities across the world
built a number of pumped-hydro facilities in the last decades, resulting in a storage
component of roughly 5% the capacity of all the European countries, 3% in the US,
and 10% in Japan. These pumped-hydro plants, and to a lesser extent compressed air
storage systems, have been used for load levelling, frequency response, and
voltage/reactive control. Likewise, storage facilities bases on other technologies such
as lead-acid batteries have been installed by a number of utilities to fulfil a variety of
functions. At a different scale, energy storage is also commonly used at the user level
to ensure reliability and power quality to customers with sensitive equipment. Another
traditional application is the electrification of off-grid networks and remote
telecommunications stations, mostly in connection with renewable sources.
The market penetration achieved by electrical energy storage to date has been heavily
constrained by its cost and the limited operational experience, resulting in high
technical and commercial risk. However the presence of storage systems is growing
fast owing to the circumstances mentioned above.
33
Electricity Storage and Wind Energy Intermittency
Final Report
Benefits of storage
Storage contributes to optimising the use of existing generation and transmission
infrastructure, reducing or deferring capital investment costs. It contributes to
integrating RE sources (and in general distributed sources) into the system, enhances
their availability and market value. The environmental benefits must be highlighted,
both in terms of reduction of the emissions from conventional power plants and
increase of RE sources penetration. Energy storage facilities can also help maintain
transmission grid stability by providing ancillary services, including black-start
capability, spinning reserve and reactive power. At the consumer level, storage
improves power quality and reliability, and can provide with capability to control or
reduce costs.
According to the European Commission (2002),
"cost-effective energy storage will be a key enabling technology for the stable
operation of a liberalised energy market, for competitive energy pricing, and
for the introduction of renewable energy sources".
The EC asserts that if energy storage systems are improved further, they can
contribute to EU policy objectives, such as meeting the Kyoto obligations to reduce
greenhouse gas emissions, lowering consumption of primary energy, creating a
sustainable supply of electricity with an increasing share of RE; and supplying lowcost reliable electricity in remote areas of Europe. Energy storage is of growing
importance as it enables the smoothening of transient and/or intermittent loads, and
downsizing of base-load capacity with consequent substantial potential for energy and
cost savings.
However, it is acknowledged that energy storage systems will have to compete within
the context of present over-capacity of power stations and power generators with short
start-up times, such as open cycle gas turbines and gas or diesel motors with the
appropriate emission controls. The EC concludes that the competitiveness of energy
storage systems against other conventional solutions is unclear.
Barriers to the deployment of electrical energy storage
Electrical energy storage involves significant investment and energy losses, which
must be weighed against the benefits and compared to other non-storage solutions.
There are a number of key barriers to a more widespread use of storage systems:
• Immaturity of some technologies and lack of operating experience. More
demonstration projects are needed to gain the customers’ confidence. Further
research and development is necessary in some aspects, such as the
implementation of power conditioning and control process for a multi-application
energy storage system.
• High initial capital costs. Technological advances and large manufacturing
volumes will bring these costs down.
• Uncertainty over the quantified benefits. This is true especially when, as
usually happens, there are multiple different benefits associated with a storage
system.
34
Final Report
Electricity Storage and Wind Energy Intermittency
• Uncertainty over the regulatory environment. The future shape of the
electricity market, not only in relation to energy trading but also to ancillary
services trading, will affect decisively the viability of electrical energy storage.
The use of storage systems for the provision of ancillary services currently
provided by the system operator will depend on the deregulatory process.
Location of storage systems
Utility-scale energy storage systems are envisaged as forming an integral part of the
future energy system. Depending on the application, they can be implemented in all
the different segments of the electric system (Fig 2.1). In a liberalised market, the
different segments of the electricity sector are being increasingly separated. Each
segment offers different potential opportunities to energy storage applications. Correct
location of the storage systems is important to maximise the benefits. Large-scale, i.e.
multiMW, centralised storage could improve generation and transmission load factors
and system stability. Smaller-scale, localised, or distributed storage could deliver
energy management and peak shaving services, as well as improving power quality
and reliability. Distributed storage would be an ideal complement to distributed
generation, especially on account of the increasing levels of RE generation.
Bulk
Generation
Transmission
Storage
Distribution
Distributed
Generation
End-use
Figure 2.1. Storage locations in the electricity supply system
One of the axioms of energy storage is that storage units should be located as close as
possible to the end consumer of electricity as possible. This is because the storage
device can improve the utilisation of all components in the network. In order to place
a storage device close to the end consumer, the device would need to be matched for
both power and energy storage capacity to the requirements of the consumer. Since
the specific capital cost increases as the system becomes smaller, the optimum
position for a storage device in the network tends to move closer to the generation
source. For this reason, Price (2000) maintains that many storage systems can, and
should be located near to substations or grid distribution points. When storage
systems are utlilised to facilitate renewable integration, the picture changes however,
since the fluctuations of the generated power are usually greater than those of the
load. As a result, the optimum location is likely to be close to the generation points,
thus maximising the capacity of the transmission and distribution lines.
35
Electricity Storage and Wind Energy Intermittency
2.2
Final Report
Applications
Applications of electrical energy storage are numerous and varied, covering a wide
spectrum, from larger scale generation and transmission related systems, to smaller
scale applications at the distribution network and the customer/end-use site. Even
though this report deals specifically with the application of storage for RE integration,
this is closely connected to other applications. Storage systems usually provide
multiple benefits, and thus it is necessary to review all their possible functions.
Interesting reviews of the applications can be found in Schoenung (2001), Herr (2002)
and Butler (2002). Ultimately, the purposes of all these applications come down to:
•
improved load management
•
provision of spinning reserve
•
transmission and distribution stabilisation and voltage regulation
•
transmissions system upgrade deferral
•
facilitating distributed generation
•
facilitating renewable energy deployment
•
end use applications
•
miscellaneous (including ancillary services)
Load management
Load management includes the traditional load levelling, a widespread application for
large energy storages, in which cheap electricity is used during off-peak hours for
charging, while discharging takes place during peak hours, providing cost savings to
the operator. In addition, load levelling can lead to more uniform load factors for the
generation, transmission and distribution systems. Although load levelling was the
first application that utilities recognized for energy storage, the differences in the
marginal cost of generation during peak and off-peak periods for many utilities is
moderate. Therefore, Butler (2002) concludes that load levelling is likely to be
provided as a secondary benefit derived from an energy storage system installed for
other applications that offer greater economic benefits. It requires energy storage
systems on the order of at least 1MW and up to hundreds ofMW, and several hours of
storage capacity (2–8 hours). For utilities without a strong seasonal demand variation,
a system used for load levelling would operate on weekdays (250 days per year).
Other types of load management are ramping and load following, in which energy
storage is used to assist generation to follow the load changes. Instantaneous match
between generation and load is necessary to maintain the generators rotating speed
and hence the frequency of the system. Storage systems serving this application
should be able to deliver on the order of 10 to 100MW to absorb and deliver power as
it fluctuates. The system would have to be able to dispatch continuously, especially
during peak load times, in frequent, shallow charging and discharging that would
occur. This service is usually provided by conventional generation.
36
Electricity Storage and Wind Energy Intermittency
Final Report
Spinning reserve
The category fast response spinning reserve corresponds to the fast responding
generation capacity that is in the state of ‘hot-stand-by’. Utilities hold it back in case
of a failure of generation units. Thus, the required power output for this application is
typically determined by the power output of the largest unit operating on-grid. The
conventional spinning reserve requires less quick response. Storage systems can
provide this application in competition with standard generation facilities.
Since the power plants that they would temporarily replace may have power ratings in
the order of 10 to 400MW, storage systems for reserve must be in this same range.
Generation outages requiring rapid reserve typically may occur about 20 to 50 times
per year. Therefore, storage facilities for rapid reserve must be able to address up to
50 significant discharges that occur randomly through the year.
Transmission and distribution stabilisation and voltage regulation
Transmission and distribution stabilisation are applications that require very high
power ratings for short durations in order to keep all components on a transmission or
distribution line in synchronous operation. This includes phase angle control, voltage
and frequency regulation.
In the event of a fault, generators may lose synchronism (difference in phase angles)
if the system is not stabilised, making the systems collapse. Energy storage devices
can stabilise the system after a fault by absorbing or delivering a power to the
generators as needed to keep them turning at the same speed. Fast action is essential
for a fast stabilisation.
Response time limitations demand an appropriate power conditioning interface design
to ensure a reliable mitigation of short-duration electrical disturbances, which can
range from a couple of cycles to two minutes. The portability of the storage systems
might be an important factor in many cases. Some applications are temporary in
nature, and Boyce (2000) points out that to transfer a storage system from site to site
can significantly increase its overall value.
With the liberalisation of the electricity market there will be an increasing need to
maintain and to improve the stability of the electrical grid. The risk of voltage
instability, being the source of failures in automatic production centres and the base of
cascading outages, will become more and more important. Many utility grids have a
limited transmission capacity with which they can properly react to transient events.
In case of fast changing load flow patterns or changes in the distribution of the loads
or power plants among the grid, the risk of voltage instability increases.
To offset the effect of the impedance in transmission lines, utilities inject reactive
power and maintain the same voltage at all locations on the line. Traditionally, fixed
and switched capacitors have provided the reactive power necessary for voltage
regulation. Storage systems deployed by transmission or distribution network
operators for other primary application can provide reactive power to the system to
augment existing capacitors and replace capacitors planned for future installation.
Energy storage system for voltage regulation should provide reactive power on the
order of 1 to 10 MVAR for several minutes, mainly during daily load peaks.
37
Electricity Storage and Wind Energy Intermittency
Final Report
Transmission upgrade deferral
When growing demand for electricity approaches the capacity of the transmission
system, utilities add new lines and transformers. Because load grows gradually, new
facilities are designed to be larger than necessary at the time of their installation, and
utilities under-use them during their first several years of operation. To defer a line or
transformer purchase, a utility can employ an energy storage system until load
demand will better use a new line or transformer.
The power requirement for this application would be on the order of 100s of kW or
several hundred MW. Butler (2002) states that the energy storage system should allow
for one to three hours of storage to provide support to the constrained transmission
facility.
Distributed generation
The growing presence of distributed sources opens a new market for storage systems,
which can assist in transient conditions of generation units such as microturbines and
diesel engines, with a slower dynamic response and thus limited capability to adjust to
load changes. In this way, storage can increase the distributed generation capacity that
can be embedded on a distribution network and avoid cost-intensive reinforcements.
A less demanding application of storage technologies in distributed generation is
peaking generation, which can also avoid reinforcement of distribution lines. Areas
with temporarily high demands, e.g. at daytime, could be equipped with storages that
supply power at peak times and are recharged through off peak hours.
These applications are often referred to as distribution capacity deferral. An energy
storage system to defer installation of new distribution capacity requires power on the
order of 10s of kW to a few MW, and must provide 1 to 3 hours of storage.
Renewable energy applications
Electrical energy storage is very promising as a means of tackling the problems
associated with the intermittency of RE sources such as wind and solar energy. The
applications will cover a wide range of power and discharge duration. With increasing
market penetration of RE these applications are more and more likely to gather
momentum within future energy systems, as conventional generation utilities ability
to even out the intermittent RE production is limited.
There are a variety of denominations in the technical literature for the use of storage
in connection with renewable applications. Butler (2002) states that some authors call
it renewable integration or renewable energy management. Schoenung (2001)
identifies only one utility-scale application under the term renewable matching,
referring to the use of storage to match renewable generation to any load profile,
making it more reliable and predictable and hence more valuable. This does not seem
to be applicable to the storage of RE at off-peak times to be delivered at peak times.
Herr (2002) however, broadens the scope of renewable matching, by referring to
applications making renewable electricity production more predictable throughout the
day and bring RE closer to demand profiles, especially providing high power outputs
38
Electricity Storage and Wind Energy Intermittency
Final Report
at peak hours. Baxter & Makansi (2002) identify four categories within RE storage:
distributed generation support, dispatchable wind, base-load wind, and off-grid
applications.
Storage systems with a longer discharge duration can cover longer mismatches (up to
several hours). In the longer term, a utility with a significant percentage of renewable
power may require storage capacity of days to ride through periods with windless
days. In Table 2.1, a number of short and long discharge renewable matching
applications are included. Both will be referred to later as renewable integration.
Indeed, a broader scope can be given to renewable integration, including short-time
applications that also contribute to tackling the problems associated with intermittent
sources.
The storage system required for either application would need to provide from 10 kW
to 100MW. According to Butler (2002), the storage system would need response time
in the fractions of seconds if transient fluctuations are to be addressed. The cycling of
the storage systems coupled with wind energy will be rather unpredictable, and could
range from one hundred to one thousand cycles per year or more.
In remote locations not connected to the grid, it may be useful to include energy
storage to minimise the generation capacity. This is especially attractive in RE-based
supplies. Renewable back-up applications should be capable of substituting RE
production when this is not available for time lengths that could go up to a week. The
power rating would depend on the corresponding power output of the RE system.
End-use applications
The primary end-use application for energy storage is power quality. Outages and
power quality phenomena are an important concern for many business sectors –a
survey estimated losses between $119 billion and $189 billion only in the US
economy. Energy storage systems are being successfully installed to provide reliable
and high quality power to sensitive loads. Transit and end-use ride-through are
applications requiring very short durations combined with very quick response times.
They cover electric transit systems with remarkable load fluctuations and customer
power services like voltage stabilisation and frequency regulation to prevent events
that can affect sensitive processing equipment and can cause data and production
losses. The demand of quality power is growing within industry and is becoming a
matter of concern also for electricity suppliers, which may also install systems at the
distribution level to improve the power quality. Uninterruptible Power Supply
(UPS) devices provide protection against electricity supply downtimes. Primarily they
prevent production losses, however, if the serving systems have very short response
times they can also be used for power quality assignments (protection against voltage
sags, power surges, frequency regulation etc). UPS systems often consist of a storage
device which usually acts during a short time until a generation set takes over.
Although the provision of UPS is usually taken on at the user level, generation
facilities can also use storage systems to remove particularly the short-term
39
Final Report
Long discharge
Short to long discharge
Fast
discharge
Electricity Storage and Wind Energy Intermittency
Application
Power rating
Discharge
duration
Storage capacity
Response
time
Transit and end-use ride-through
< 1MW
seconds
~2 kWh
< ¼ cycle
Transmission
stabilisation
up to 100's MVA
seconds
20 – 50 kVAh
< 1/4 cycle
Voltage regulation
up to 10 MVAR
minutes
250 – 2,500 kVArh
< 1/4 cycle
Transmission
Fast response spinning reserve
10 – 100MW
< 30 m
5,000 – 500,000 kWh
<3s
Generation
Conventional spinning reserve
10 – 100MW
< 30 m
5,000 – 500,000 kWh
< 10 min
Generation
Uninterruptible power supply
< 2MW
~2h
100 – 4,000 kWh
seconds
End-use
End-use & transmission peak shaving
< 5MW
1–3h
1,000 – 150,000 kWh
seconds
End-use
Distribution
Transmission upgrade deferral
up to 100'sMW
1–3h
1,000 – 500,000 kWh
seconds
Transmission
Renewable matching (short discharge)
< 100MW
min – 1 h
10 – 100,000 kWh
< 1 cycle
Generation
Renewable matching (long discharge)
< 100MW
1 h – 10 h
1,000 – 100,000 kWh
seconds
Generation
Load levelling
100'sMW
6 – 10 h
100 – 10,000MWh
minutes
Generation
Load following
10 – 100'sMW
several hours
10 – 1,000MWh
< cycle
Generation
Distribution
Emergency back-up
< 1MW
24 h
24MWh
seconds
End-use
Renewables back-up
100 kW – 1MW
days
20 –200MWh
sec – mins
Generation
End-use
&
distribution
System location
End-use
&
Distribution
Transmission &
Distribution
&
&
&
Table 2.1Applications of storage systems with different discharge times
40
Electricity Storage and Wind Energy Intermittency
Final Report
fluctuations from their supply. The attractiveness of the investment will depend on
any penalties imposed on generating units that fail to provide a quality supply.
There are other customer uses such as end-use peak shaving that can avoid demand
charges by reducing demand peaks. Emergency back-up at customer site requires
power ratings of approximately one MW for durations up to one day. Presently, most
of these applications are served by reciprocating engines.
Miscellany
The provision of ancillary services by storage systems can also include black start
capability, which consists of the supply of electricity for the start up of generators
after a network failure. It is usually performed by relatively expensive diesel engines.
Some storage options also need an auxiliary electricity supply, but several can start
without an electricity source. Once again this service can be provided in addition to
other applications listed
A number of different lists and different descriptions of the storage applications can
also be found in the literature. Some authors include deferment of new capital
equipment as a separate category. This is in fact simply an aggregation of some of the
applications already quoted, which can be performed by conventional equipment such
as peaking plant, new lines, substations, etc. The installation of storage systems on the
transmission and distribution grid in order to expand the grid capacity, decouple
generation and load, and thus reduce congestion, has already been included as a
separate application (transmission upgrade deferral). The use of storage systems to
improve transmission stability also reduces or defers the need for transmission
upgrades. Likewise, storage units for load levelling, spinning reserve or peak shaving
delay the need for new generation capacity.
Other authors refer also to the improvement of power plant efficiency as a category,
but this is rather a driving force for applications such as load management and
spinning reserve.
Environmental benefits is also sometimes quoted as an application, but it is rather a
consequence derived of the application of storage systems as spinning reserve, peak
shaving and others, which results in the cut of emissions that conventional
technologies cause. Energy storage can enhance the environmental performance of a
network in a number of ways:
Conventional generating units used to provide spinning reserve and other
ancillary services can be replaced by energy storage.
Generators which operate best at constant load can be combined to provide
ramping and peaking duties.
Grid upgrades can be avoided.
System control issues arising from intermittent RE sources can be mitigated,
thus increasing the proportion of renewable generation that the system can
absorb.
41
Final Report
Electricity Storage and Wind Energy Intermittency
2.3
Technical requirements of storage applications
The many applications can be characterised by their technical requirements, i.e. power
level, energy storage capacity, and response time. There are other possible
requirements which may be relevant in some applications and in some cases, such as
portability and limited footprint. This may be an important issue for transmission
capacity deferral purposes, but will be irrelevant in applications involving large
energy capacity.
The discharge time, which is basically determined by the power/energy ratio,
influences to a great extent the design of the storage system and the technology
selection. Whereas some applications demand an energy discharge burst lasting only a
few seconds or even less (power quality and stability), others require the energy to be
available for long periods of time, up to a few hours or even days (load management)
as illustrated in figure 2.2.
Power quality
(cycles –seconds)
Reliability &
Productivity
Load following
(minutes)
HIGH
POWER
Energy management
(hours)
HIGH
ENERGY
Profitability &
Renewables
Figure 2.32 Discharge timeframes of different storage applications
The power/energy ratio will dictate to a great extent the design of the storage system
and the technology selection as figure 2.3 illustrates.
1.E+04
Energy (MWh)
1.E+02
Levelling
RE back-up
Back-up
1.E+00
RE
matching
UPS Peak
Discharge
time
Spinning
1s
1m
1h
10 h
1.E-02
1.E-04
Transmission stability
Power quality & reliability
1.E-06
0.01
0.1
1
10
100
Power rating (MW)
Figure 2.3 Energy and power ratings of storage applications
42
Electricity Storage and Wind Energy Intermittency
Final Report
Response time is another relevant factor, which becomes critical in stabilisation and
power quality applications. Table 2.1, based mainly on Schoenung (2001), shows the
requirements of the different storage applications. Since storage systems are likely to
perform more than one function to achieve a better economic viability, the
requirements for multifunctional systems will obviously be more stringent.
2.4
Market potential of storage applications
Electrical energy storage will clearly only penetrate the market if it proves the most
cost-effective solution. The decision to use an energy storage system depends both
upon the requirements of the application and the cost of competing solutions.
Power quality applications currently show the best cost-benefit quotient. As
production lines become more and more automated, small instabilities are proving to
be increasingly costly to industry. Some studies indicate very significant production
losses in sectors such as the semiconductor and pharmaceutical industries as a
consequence of short interruptions. As a result, certain applications could be regarded
as imperative, e.g. the maintenance of power quality, voltage stability, generation
adequacy, and so forth. Storage systems for power quality are already being used at
the customer level in many factories where power quality is a critical issue, however,
in today's ‘digital economy’, customers demands of a more reliable and high-quality
supply will progressively become more intensive.
Baxter (2002) points to a US-based study showing that transmission applications
would have the greatest economic impact in US ($129.9 billion). End-use power
quality and reliability would have an impact of $31.2 billion, whereas generationbased applications (load levelling, load following, spinning reserve) would have a
more modest impact of $10.6 billion in the US market.
There is controversy over the viability of storage in the integration of renewables.
Boyes (2000) points out that some studies suggest that the operating savings
associated with the implementing of storage systems in connection with intermittent
renewable energies can be four to six times greater than those from adding storage to
a utility system without RE.
The growth of intermittent RE sources will enlarge the potential for energy storage far
beyond the traditional levels. The competitiveness of storage against non-storage
solutions will be influenced by a number of factors, such as the scale of renewable
penetration in an electricity system, which could make other options unfeasible. The
possible internalisation of environmental effects of conventional generating
technology could also tip the balance in favour of storage.
2.5
Renewable energy storage applications
The focus of this report is the integration of wind energy and energy storage, or, how
storage technologies can overcome the problems arising from the intermittent nature
of wind energy.
43
Electricity Storage and Wind Energy Intermittency
Final Report
The two categories of storage applications identified previously as associated with
renewable integration (matching and back-up) are merely adhering to a convention.
Storage systems can support intermittent RE in many issues related to its integration
in the network. In the classification above these applications have been distinguished
from renewable integration. In short, whereas in a system without intermittent RE
sources the only uncertainty is the demand (apart from failures risks) the presence of
fluctuant generation adds a new source of uncertainty, which becomes more dominant
as the renewable penetration increases. The possible benefits are thus widespread and
may be grouped under the headings energy trading and network services.:
Energy trading
•
•
•
Load levelling, benefiting of the on-peak/off-peak price differential
Avoidance of penalties on power exchanges due to predictable electricity
deliverance
Avoidance of wind energy curtailments
Network services
•
•
•
•
•
•
Voltage regulation, reactive compensation.
Transmission and distribution stability
Frequency control
Avoidance or deferral of transmission and distribution upgrading
Provision of spinning reserve
Black start capability
These applications can also be condensed into three: time shifting generation, control,
and reserve. The different applications are linked to the durations of wind variations.
Short-time fluctuations (over a few seconds) demand services such as stabilisation
and frequency regulation. Longer fluctuations (hourly, daily or longer) are associated
with energy trading applications. A combination of short and long-time energy
storage enhances the economic viability of renewables according to Collinson (1999).
The provision of reserve has to do with the limited capability of intermittent sources
to displace generation plant capacity. System operators will need to schedule more
reserve as the amount of RE on an electricity system increases and adds to the
uncertainty in balancing supply and demand. This is a key issue in the context of the
Irish electricity market as the provision of reserve will be carried out under a bidding
system and charging for reserve will be carried out on the basis of the causer pays
principle. The impacts of increased wind penetration on operating reserve
requirements is currently the focus of an SEI funded study being led by UCD.
Although the provision of reserve alone with storage systems is unlikely to prove
viable, systems serving other applications can also increase the capacity credit of
intermittent sources and reduce their loss of value. A flexible and fast-response
system is ideal for this purpose. Storage allows dispatch to follow load curves or to be
held in reserve.
Ingram (2000) asserts that storage can optimise the transmission of renewable
electricity, and hence maximise the use of the existing grid capacity. Wind farms are
44
Electricity Storage and Wind Energy Intermittency
Final Report
often located at the end of long transmission lines with little or no available capacity
during high load hours. Even if sufficient transmission capacity is available, in some
markets renewable generators face the cost of grid-based transmission rights and
losses. Purchases of transmission based on peak output have a worse impact on
fluctuant sources owing to the low capacity factors.
Practically all energy storage facilities are expected to perform a number of support
functions. Storage facilities providing load levelling can also maintain the stability of
the system and provide reserve capacity.
Scale and location of the storage systems
The selection of suitable storage technologies and the sizing of the systems will
depend on the functions which will be fulfilled, as well as the location of the system
in the network. In this context, it is useful to draw distinctions between the following
categories :•
large-scale generation
•
distributed generation
•
off-grid
The fast growth of intermittent renewable sources in many countries, like Ireland,
prompts expectations for the potential of large-scale storage systems, which would
participate in the wholesale market. Large storage system can play a key part in a
strategy to minimise the total cost of power delivered. The power rating could range
from some hundred kilowatts to over a hundred megawatts.
Storage can help in the integration of RE in the distribution network, especially in
weak grids, where the capacity to accommodate intermittent sources, especially wind,
may be strongly constrained. Local loads can be partially met with stored energy at
times when the RE source is unavailable. In short, the installation of distributed RE in
connection with storage can defer the upgrade of distribution lines according to
ORNL (1997). Connected to a particular wind farm, storage can help to meet the
connection regulations and avoid connection charges. The size of storage systems
connected to a distribution network will be smaller than in the wholesale market,
depending on the amount of embedded renewable generation.
The size of off-grid systems ranges from small residential or telecommunication
systems to medium size isolated networks. The presence of intermittent sources in an
isolated system makes storage much more important than in a highly interconnected
network. Such is the case of the island of Crete, the largest off-grid system in the
world. 624MW of wind energy are predicted in Crete for 2010, as well as 213MW of
pumped storage units within the Action Plan for Large Scale Deployment of
Renewables. Ireland has only a low-capacity link with other electricity systems, and
thereby it is in a way similar to the Crete system, where significant wind curtailment
is necessary at times.
Off-grid applications are usually to be found in the electrification of small systems.
Many off-grid locations are in environmentally sensitive areas. Renewable energy and
storage can obviate the need for the fossil technology with its accompanying supply
45
Electricity Storage and Wind Energy Intermittency
Final Report
infrastructure. This usually requires large storage capacities to offset long periods of
renewable unavailability. Therefore, fossil fuel support is usually necessary; storage
can however, achieve great savings in the fuel consumption and reduce largely the
cycling of the generators. The ultimate benefits are the same as in utility-scale storage
–efficiency of the system and cost reduction–. There is very little market for off-grid
systems in Ireland, probably limited to some very small applications, and therefore
they fall beyond the scope of this study. It should be noted however, that the
development of expertise and technology in this area could benefit Ireland from the
perspective of export potential.
2.6
Research and investment in electrical energy storage
The growing interest in electrical energy storage has prompted an increase in research
and investments worldwide.
European Union research
The European Commission’s 5th Framework Programme of the included a target
action on energy storage, which provided a strategic focus on the medium to long
term needs for research on storage.
In the 6th Framework Programme, stationary energy storage is included among the
research activities having an impact in the short to medium term. It is acknowledge
that short-term research on the large-scale integration of RE sources into energy
supplies is needed in support of the EU’s commitments to increase the percentage of
renewables in its supply mix. Electricity storage is among the areas in which the EU
envisages support in particular, including ‘advanced batteries, hydrogen, and other
electricity storage devices for balancing variations in renewable electricity supply’.
Proposals for Integrated Projects have been invited for the topic Advanced energy
storage systems for RES, with the objective is to develop technologies and systems
for the storage of electricity for grid-connected applications enabling the increased
penetration of renewable and distributed generation of electricity in new distributed
electricity networks. R&D should also consider the analysis of storage system
performance (in terms of lifetime, system lifetime cost, reliability, safety and
recyclability of materials), the benchmarking of technologies and pre-normative
research. One of the Specific Targeted Research Areas is: Electricity – innovative
energy storage technologies for grid-connected applications (new concepts for
energy storage technologies, where applicable exploiting the synergies with transport
applications).
A current project funded by the EC is the Investire Network (Investigation on
storage technologies for intermittent renewable strategies). The objectives are to
review and assess existing storage technologies in the context of renewable energy
applications to facilitate exchange of information. In Europe, there are few countries
in which actors are capable of developing to a significant level more than two storage
technologies, when so many technologies are claimed to be potential candidates to the
various renewable energy applications. The project seeks to exchange R&D
information, disseminate experiences and increase the markets, for example HES
46
Electricity Storage and Wind Energy Intermittency
Final Report
from Scandinavia, FES from UK, short-term storage for wind systems from Denmark,
and so on.
International Energy Agency
The R&D programme Efficient Energy End-Use Technologies of the IEA contains 14
different Implementing Agreements of which one is ‘Energy Conservation through
Energy Storage’ (ECES IA). According to the IEA (2003a)
“The overall objective is to develop and demonstrate various energy storage
technologies for applications within a variety of energy systems and to
encourage their use as a standard design option. Energy storage technologies
can improve the utilisation of renewable energies, in particular solar and wind
and the greater utilisation of waste heat energy storage technologies should be
implemented in all countries with significant energy storage market potential’’ .
At present, the ECES IA contains 17 different Annexes, some of which have been
terminated. Most of the them are related to thermal energy storage. Relevant to
electrical energy storage are the Annex IX, entitled ‘Electrical Energy Storage
Technologies for Utility Network Optimisation’, and the recently proposed Annex
XV, entitled ‘Electrical Energy Storage and the Integration of Renewables’.
Annex IX had the task of examining the potential role of electrical storage
technologies in optimising electricity supply and utilisation. It also sought to identify
barriers to widespread adoption of electrical energy storage technology.
The project has produced several reports, including a case study report of energy
storage systems and a project definition report, which defines the framework for two
potential demonstration projects (one for a power quality application and one for a
utility-scale bulk storage project). A computer model for evaluating power quality
applications has been developed and a requirement specification has been produced
for the definition of a network applications model.
Annex XV is a natural development borne out of Annex 9. The aim is to develop a
firm understanding of the technical issues and commercial implications of applying
electrical energy storage technologies to the integration of RE and to develop
awareness of the capabilities and uses of existing and developing energy storage
systems as applied to RE (IEA. 2003). One objective is to move storage systems
towards commercial market implementation, via the mechanism of technological and
applications demonstrations.
The basic proposed activities focus on:
•
the need for storage from a renewables perspective
•
modelling of network/renewables/storage interaction
•
implementation strategies for storage-based solutions
•
the costs of storage
•
the benefits of storage
•
alternatives to storage
47
Electricity Storage and Wind Energy Intermittency
Final Report
The conferences on Electrical Energy Storage Applications and Technologies
(EESAT) provide an opportunity for dissemination of results from the Annex IX
activities and to discuss issues related to the market, the applications and the
technologies. So far there EESAT conferences have been held, in the years 1998,
2000 and 2002. Further information can be obtained from www.iea-eces.org/
United States
US have been leaders in the research and development of electrical energy storage.
The main driver is the Energy Storage Systems Program of the US Department of
Energy. This programme, conducted by Sandia National Laboratories involves
systems integration, component development, prototype testing and systems analysis.
Many projects are performed in collaboration with private sector organizations.
www.sandia.gov/ess/
www.eere.energy.gov/
www.eere.energy.gov/EE/power_energy_storage.html
Other useful links
The Electricity Storage Association (ESA) is an industry trade organization founded
by eight electric utilities in 1990 that perceived a viable role for energy storage in
electric power applications. Originally focused on battery energy storage, the
organization was founded as an informal association as the Utility Battery Group, and
later incorporated as the Energy Storage Association. The ESA is now a membership
trade association that has the mission of fostering development and commercialisation
of competitive and reliable energy storage delivery systems for use by electricity
suppliers and their customers.
www.electricitystorage.org/
EA technology is a British utility consultant with considerable expertise on energy
storage, and participated in the Annex IX of the ECES IA.
www.eatechnology.com/utilities_business/storage.cfm
The American consultant Zaiginger Engineering Company, Inc. has participated in
many studies relating to distributed generation, renewable energies and storage
systems.
www.zecoconsulting.com/distributed_gen_&_storage.htm
US Energy Storage Council
http://www.energystoragecouncil.org/
48
Final Report
Electricity Storage and Wind Energy Intermittency
Chapter 3 STORAGE TECHNOLOGIES
Storage systems generally comprise three key elements, namely Storage Subsystems,
Power Conversion Systems (PCS) and Balance of Plant Systems (BOP) as illustrated
in figure 3.1. Depending on the storage system, certain elements within the scheme
may be unnecessary, e.g. pumped hydro and compressed air energy storage do not
need a rectifier and inverter, as pumps and compressors operate using alternative
current (AC).
Grid/Load interface
Power Conversion System (PCS)
Transformer
&
Rectifier
Transformer
&
Inverter
Control unit
Storage subsystem
Converting devices
Storage medium
Services
Balance of Plant (BOP)
Building
Heating/Ventilation/Air Conditioning
Figure 3.1 Scheme of a storage system
There is a wide range of energy storage technologies at utility scale that are at various
stages of development. Each technology has different features which make it more or
less desirable for the various applications. Table 3.1 provides an overview of possible
selection criteria.
The relevance of the different features varies largely depending on the application
which is going to be served. Fundamental criteria for any technology will be the
power capacity (including the reactive power capacity for some purposes), the
energy capacity/discharge time, and the reaction time. Some applications, like grid
support, require discharges to commence less than a second after beginning; others,
like power sales, can be scheduled allowing for a reaction time of a few minutes.
49
Final Report
Electricity Storage and Wind Energy Intermittency
Design
Operating
Financial
Others
• Power rating
• Storage capacity/
discharge duration
• Response time
• Energy density per
unit area (footprint)
• Energy density per
unit volume and weight
• Maturity of
technology
• Reliability
• Modularity
• Siting requirements
• Portability
• Synergies with other
energy applications
• Overall cycle
efficiency
• Lifetime/maximum
number of chargedischarge cycles
• Parasitic losses
• Capital cost per
energy stored
• Capital cost per
power rating
• Fixed O&M cost
• Variable O&M cost
• Replacement cost
• Disposal cost
• Commercial risk
• Health and safety
aspects
• Environmental
impacts
• Synergies with other
sectors
Table 3.1. Criteria for the selection of a storage technology
Ideally energy storage technologies should:
•
be low capital, operating and maintenance cost
•
have a long lifetime
•
be flexible in operation
•
have a high efficiency
•
have a fast response
•
be environmentally sustainable
There is a notable absence of available detailed technical information about storage
technologies. This is surprising, given the growing need and opportunites for storage
technologies. Somewhat superficial reviews can be found in Gandy (2000);
Schoenung (2001); Swaminathan (1997); Ter-Gazarian (1994) and Herr (2002).
Electricity storage systems can be technically categorized by their inherent physical
principles into mechanical, electromagnetic and electrochemical storage devices.
Mechanical
Electromagnetic
Electrochemical
Pumped Hydro
Super-Capacitors
Batteries
Compressed Air
Super-Conducting
Magnets
Flow Batteries
Flywheel
Hydrogen
Figure 3.2 Storage technology categories
50
Electricity Storage and Wind Energy Intermittency
3.1
Final Report
Pumped hydro
Shoenung (2001) acknowledges the role of Pumped Hydro Energy Storage (PHES) as
the most widespread energy storage system currently in use on power networks,
operating at power rating up to 4,000MW and capacities up to 15 GWh. PHES uses
the potential energy of water, swapped by pumps (charging mode) and turbines
(discharge mode) between two reservoirs located at different altitudes. Currently, the
overall efficiency is in the 70-85% range although variable speed machines are now
being used to improve this. The efficiency is limited by the efficiency of the deployed
pumps and turbines (neglecting friction losses in pipes and water losses due to
evaporation).
Plants are characterized by long construction times and high capital costs. One of the
major problems related to building new plants is of an ecological/environmental
nature. At least two water reservoirs are needed. Some high dam hydro plants have a
storage capability and can be dispatched as a pumped hydro. A relatively new concept
of pumped hydro employs a lower reservoir buried deep in the ground. A good
example of underground pumped storage is the Dinorwig plant in UK, commissioned
in 1982, which includes Europe’s largest man-made cavern under the hills of North
Wales. Open sea can also be used as the lower reservoir –a seawater pumped hydro
plant was first built in Japan in 1999.
Pumped hydro facilities are available at almost any scale with discharge times ranging
from several hours to a few days. PHES can be designed for fast loading and ramping,
allowing frequent and rapid (<15 sec) changes among the pumping, generating and
stand-by spinning modes (Gordon, 1995). The Dinorwig plant can go from 0 to
1890MW (full capacity) in only 16 seconds. PHES is best suited to load levelling,
storing energy during off-peak hours for use during peak hours, and spinning reserve.
They can provide energy to meet peak demands, and in the pumping mode, they serve
as the source of load for base-load during off-peak periods, helping to avoid cycling
these units and improving their operating efficiency. PHES systems can provide other
benefits, including black start capability (they can begin generating without an
external power source) and frequency regulation.
There is over 90 GW of pumped storage in operation world wide in nearly 300 plants,
which is about 3 % of global generation capacity. In 1998 10% of Japan’s total
instantaneous energy requirement came from pumped hydro (Tanaka, 1998). Table
3.2 contains some of the most representative pumped-hydro plans in the world. The
292MW Turlough Hill Pumped Storage station (representing approx 5% of installed
capacity) is the only bulk electrical energy storage facility in Ireland. Its construction
was completed in 1974 and involved the construction of a huge cavern in the heart of
the mountain, in which the generation plant and controls are housed. A pumped
storage system allows for the use of excess electricity capacity during non-peak hours
to pump water from the lower to the upper lake at Turlough Hill and then the release
of the water in the reverse direction to create electricity in times of maximum
demand.
Unit prices for pump/turbines have levelled out as the technology has matured. Thus,
costs are typically around $600/kW. Reservoir costs can vary from almost nothing to
51
Final Report
Electricity Storage and Wind Energy Intermittency
more than $20/kWh according to Gordon (1995). Schoenung places that cost at
$12/kWh.
Developers / Suppliers: MWH, GE Hydro, First Hydro Company.
Country
Location
Date
China
Germany
Japan
Taiwan
UK
USA
USA
USA
Tianhuangpin
Goldisthal
Kazunogowa
Mingtan
Dinorwig
Northfield
Bad Creek
Bath County
2001
2002
2001
1994
1994
1973
1991
1985
Max Power
(MW)
1800
1060
1600
1620
1890
1080
1065
2700
Hours of
discharge
8.2
5
10
24
11
Plant cost
$1,080 M
$700 M
$3,200 M
$1,338 M
$310 M
$685 M
$652 M
$1,850 M
Table 3.2 Some of the largest pumped-hydro facilities in the world
3.2
Compressed air
Compressed Air Energy Storage (CAES) systems are used, like PHES, for storing
large amounts of energy, though they are much less employed worldwide. The
electricity is stored by compressing air via electrical compressors in huge storage
facilities, mostly situated underground in caverns created inside appropriate salt rocks,
abandoned hard-rock mines, or natural aquifers, as discussed in Schoenung (2001).
Recovery takes place by expanding the compressed air through a turbine, but the
operating units worldwide incorporate combustion prior to turbine expansion in order
to increase the overall efficiency of the system. Hence CAES can be regarded as
peaking gas turbine power plants, but with a higher efficiency, thanks to the
decoupling of compressor and turbine, and much lower overall cost. The savings
come from the fact that, unlike conventional gas turbines that consume about 66% of
their input fuel to compress air at the time of generation, CAES pre-compresses air
using the low cost electricity from the power grid at off-peak times and utilizes that
energy later along with some gas fuel to generate electricity as needed. The
electricity/fuel ratio is an important design criterion for CAES plants.
Since CAES uses two energy sources – natural gas and electricity – it is difficult to
specify efficiency in a meaningful way. Based on the efficiency of compression and
expansion, Herr (2002) gives an efficiency of 64% for large systems.
Size limitations are driven mostly by the size of the gas turbines available. Currently,
units as small as 20 kW are available. Large size units are limited by the reservoir
size, as well as the grid capacity.
CAES offers an alternative to PHES for the storage of a large amount of power, most
usefully for load levelling. It can also provide ancillary services, including reactive
power. CAES plants can ramp faster than simple-cycle gas-fired plants because they
are not restrained by compression requirements. Zink (1997) points to studies
concluding that CAES is competitive with combustion turbines and combined-cycle
units, even without attributing some of the unique benefits of energy storage.
52
Electricity Storage and Wind Energy Intermittency
Final Report
However, few projects have been successfully completed globally, and so it remains a
technology of some potential but little experience. The site-specific nature, coupled
with the modest demand for long-duration storage, has limited the market entry of
CAES. Owing to the limited operational experience, the technical risk is considered
high by many utilities according to Gordon (1995). Price (2000) points out that the
recently announced proposals for micro CAES using small gas turbines and pipelines
as air receivers may reverse this trend. Micro CAES could be conveniently located
near to load centres and become a useful distributed resource.
The first commercial CAES system was a 290MW unit built by ABB in Huntorf,
Germany in 1978 (Figure. 3.3). This plant, now decommissioned, was operational for
10 years with 90% availability and 99% reliability according to Breeze (1998). The
storage reservoir was a 300,000 m3 underground cavity in a natural salt deposit, where
air was stored at 70 bar. The system was charged over an eight-hour period, and could
delivered 300MW for 2 hours.
Figure 3.3
290MW CAES plant in Huntorf
The second commercial unit was a 110MW unit built by Dresser-Rand in McIntosh,
Alabama in 1991. The construction took 30 months and cost $65M (about $591/kW).
Semadeni (2003) reports that the plant has since generated over 55 GWh during peak
demand periods. It comes on line within 14 minutes and can supply the nominal
power for 26 h according to Price (2000).
The third commercial CAES plant, the largest ever (and larger than any energy
storage plant in US, including pumped hydro), is a 2,700MW plant that is being
developed in Norton, Ohio, by CAES Development Company. Van der Linden (2002)
explains that this 9-unit plant will compress air to 104 bar in an existing limestone
mine some 670 m under ground, with a capacity of 9.5 million m3. Also in progress is
the 540MW facility in Markham, Texas, being developed by Ridge Energy Storage.
There are additional CAES plants built or planned at Sesta in Italy (25MW), in Japan
(35MW, 6 h), Israel (300MW) and Russia (1050MW).
The features and limitations of CAES are similar to those of PHES. The need for
geologically suitable locations for underground storage acts as a significant constraint
to the deployment of this technology. Salt caverns are created by drilling a
conventional well to pump fresh water into a salt dome or bedded salt formation. The
salt dissolves until the water is saturated, and the resulting salt water is returned to the
surface. This process continues until a cavern of the desired volume and shape is
53
Electricity Storage and Wind Energy Intermittency
Final Report
created. It can take about 1.5 to 2 years to create such a cavern. Detailed studies of
underground storage caverns are essential before excavation and are very expensive.
Taylor (1999) notes that hard-rock caverns are more costly to mine (60% higher) than
salt-caverns for CAES purposes. Smaller on-site plants may be built using
aboveground man-made reservoirs, possibly posing special safety or permitting
challenges. Another possibility is the use of fabricated high-pressure tanks. Because
of the expense of such tanks, only several hours worth of storage has been proposed
for this concept according to Nakhamkin (1999).
In contrast to other storage technologies CAES is dependent on supplies of primary
fuel in addition to an electrical supply. Air emissions (from combustion of gas) and
most safety issues are very similar to other gas turbine-based generation plants.
Ridge Energy designs standard compression train blocks of 100MW each and
standard generation blocks of 135MW. In generation mode, the plant can start up
from 0 to 100% in less than 10 minutes. A normal ramp up from 10 to 100% load is 4
minutes, while in emergency it can be done in 2 minutes. Ramping from 50% to
100% can be accomplished in less than 15 seconds. As for the compression, the full
load is reached in less than 10 minutes, and the 50% - 100% ramp in less than 10
seconds. They are capable of black start.
Schoenung (2001) and Gordon (1995) project capital costs to range between $425 and
$480/kW for advanced designs if expected commercialisation occurs and expected
experience is gained. Energy related costs are estimated between $3/kWh by
Schoenung (2001) and $10/kWh by Gordon (1995). Costs depend largely on special
requirements related to geologic reservoirs. The O&M costs (excluding fuel) will also
be heavily affected by the reservoir characteristics.
Developers / Suppliers: CAES Development Company, Ridge Energy Storage,
Dresser-Rand.
3.3
Flywheels
Kinetic energy may also be used to store energy in the form of the inertia of a
flywheel. Flywheels have been used for centuries to flatten intermittent input in such
applications as windmills. Nowadays all reciprocating engines contain flywheels to
smooth the pulsed output of the pistons and provide stable power. With the advent of
advanced composite materials with high tensile strength, and the development of
stable magnetically suspended bearings, flywheels may now be made with
significantly higher operational speeds. This in turn opens a new field of opportunities
for Flywheel Energy Storage (FES).
Most modern flywheel energy storage systems consist of a massive rotating cylinder
(comprised of a rim attached to a shaft) that is substantially supported on a stator by
magnetically levitated bearings that eliminate bearing wear and increase system life.
To maintain efficiency, the flywheel system is operated in a low vacuum environment
to reduce friction. The flywheel is connected to a motor/generator mounted onto the
stator that, through some power electronics, interact with the grid. The basic diagram
of a FES system is shown in Figure. 3.4. Some of the key features of flywheels are
54
Electricity Storage and Wind Energy Intermittency
Final Report
rapid dynamic response, little maintenance, long life (20 years or 10s of thousands of
deep cycles) and environmentally inert material.
The design of the flywheel itself has been the subject of much research. In order to
maximise the energy stored, mass, radius and angular velocity of the flywheel must be
increased, but this poses challenges to the materials strength. In general, flywheels
can be divided into low-speed and high-speed units. The former are made of steel, and
the latter, which reach speeds up to 100,000 rpm, of low-density composite materials
such as fibre-reinforced carbon, aramid and glass fibres. The choice is based on the
system cost, weight, size, and performance.
Flywheels storage systems are particularly suitable for power quality control. They
can provide ride-through power for the majority of power disturbances, such as
voltage sags and surges, and can bridge the gap between a power outage and the time
required to switch to long-term storage or generator power with excellent load
following characteristics. In comparison with lead-acid batteries providing ridethrough and UPS, flywheels have a longer lifespan, lower maintenance, faster
charge/discharge, take up relatively little space and pose no environmental hazards. In
these applications, the need for a rapid load following, very frequent cycling and high
power draws negatively impact battery life.
Flywheels are therefore contemplated now only for a range of short-term applications
up to a size of severalMW, although not only for quality power and UPS. Despite the
moderate storage capacity, some interest has been shown for the use of FES in
renewable generation applications, basically to smooth out the power output and thus
help improve power quality, and in stand alone systems in conjunction with batteries.
Butler (2002) asserts that steel rotor FES have limited promise for the entire array of
applications but is well suited to hybrid FES/battery power quality applications. He
also states that composite-rotor FES has potential for broader applicability but will
require significant development to compete with other, more mature technologies. A
number of technical hurdles will need to be overcome and the level of the technology
will need to mature before flywheels become more widespread. High initial costs
have slowed adoption of these units, although life-cycle costs can be already lower
than for battery systems, especially in demanding operations. Research on flywheels
focuses on improvements in the materials and manufacturing processes to achieve
long-term mechanical stability, better low-loss bearings and cost reduction.
Beacon has developed a 250 kW/25 kWh flywheel. A matrix containing 10 units
could deliver 2.5MW during 5min, or 0.5MW during 30min. The most powerful
referenced flywheel is from the University of Texas, which is able to discharge 3MW
for 2½ minutes according to Taylor (1999).
Headifen (1994) examined flywheels used for load-levelling in conjunction with 300
kW wind turbines (to smooth out power variations) gave a detailed cost breakdown
for a 300kW/277kWh flywheel, resulting in a total installed cost of $220,000 or about
$800 per kWh. The manufacturer Beacon, however, expects composite flywheels to
break even with steel machines at 2-6 kWh, and is aiming at $500/kWh in composite
units of 25 kWh. Other developers estimate long-term costs as low as $200/kWh
according to Akhil (1997). Schoenung (2001) gives a cost estimate of $200-300/kWh
55
Electricity Storage and Wind Energy Intermittency
Final Report
for low-speed flywheels, but, surprisingly, the cost estimated for high-speed systems
soars to 25,000/kWh. This wide range of cost estimates is typical for a technology in
its early stages of development. In any case, composite flywheels are progressively
becoming competitively priced, especially in applications requiring high power rating.
Costs are expected to drop with increasing energy, but so far, only modest sizes have
been reached. It must be borne in mind that the long life span and low O&M costs
make the lifetime cost much more affordable. Although flywheel banks have been
proposed to reach higher energy capacities, this solution does not avail of economyof-scale effects. The operating costs of large units, dominated by vacuum pumping,
are expected to be very low according to Gordon (1995).
As Manwell, McGowan & Rogers (2003) point out, some variable speed pitch
regulated wind turbines use rotor speed controls to improve power quality in wind
turbines. During gusts, the generator power is maintained at a constant while the rotor
speed increases. The increased energy in the wind is stored as kinetic energy in the
rotor. If the wind speed drops, the reduced aerodynamic torque results in a
deceleration of the rotor speed while the generator power is kept constant. This
principle of energy storage as kinetic energy in the rotor is similar to the principal
underpinning flywheel energy storage.
Developers / Suppliers: Beacon, Active Power, AFS Trinity Power,
Urenco Power Technologies, Flywheel Energy Systems, ASPES AG,
Pentadyne Power Corporation, Piller, Regenerative Power and Motion.
Figure 3.4 Basic diagram of a FES system
3.4
Super-capacitors
Capacitors store energy by way of separating the charge onto two facing plates. They
are widely used in electronic devices for power smoothing after rectifying. Typically,
these applications require very small energy amounts. In order to increase the energy
density, the so-called ‘Super-Capacitors’ (or even ‘Ultra-capacitors’, if their
capacitance exceeds 1000F) have been developed. They use polarized liquid layers at
the interface between a conducting ionic electrolyte and a conducting electrode, which
increases the capacitance. The bipolar configuration lends itself to versatility in
56
Electricity Storage and Wind Energy Intermittency
Final Report
connecting individual cells in series or parallel. Energy densities of 20 to 70 MJ/m3
are usually reached. The efficiency is approximately 95% (Gordon, 1995).
Super-Capacitors Energy Storage (SCES) offers extremely fast charge and discharge
capability, albeit with a lower energy density than conventional batteries can provide
and can be cycled tens of thousands of times without degradation.
The interest in automotive applications and telecommunications has stimulated the
development of the technology. SCES will probably be linked to batteries in these
realisations. By combining a supercapacitor with a battery-based uninterruptible
power supply system, the lifetime of the battery is extended. The batteries would
provide power only during the longer interruptions, reducing their cycling duty on the
battery. Small supercapacitors are commercially available to extend battery life in
electronic equipment, but large supercapacitors are still in development.
SCES is still at an early stage of development, as an energy storage technology for
electric utility applications. Small-scale power quality (<250 kW) is considered to be
its most promising utility use. The units under development target applications where
pulse power is needed for duration in the millisecond to second time range, mainly to
by-pass voltage sags. Supercapacitors with discharge times up to 1 minute are
feasible.
In any event, their use for large energy storage is not contemplated in the medium
term. Since the number of capacitors needed is directly proportional to the stored
energy, there is no economy of scale. This, together with the relatively low energy
densities, makes them little attractive for large energy (or long duration) applications.
Gordon (1995) reports that capital costs of SCES were estimated by Pinnacle
Research Institute at $12,960/kWh, with O&M 5% of the capital cost per year.
Nevertheless, Schoenung (2001) estimates the energy-related costs at $82,000/kWh.
Further research has to concentrate on cost, manufacturing process and the lowering
of internal resistance.
Developers: SAFT, NessCap, ESMA, Maxwell, ABB, ELIT.
3.5
Superconducting magnets
In a Superconducting Magnet Energy Storage (SMES) device, a coil of
superconducting wire allows a DC current to flow through it with virtually no loss.
The current creates a magnetic field that stores the energy. On discharge, special
switches tap the circulating current and release it to serve a load. For setting the coil
in state of superconducting, it has to be cooled down either to 4.2°K (low-temperature
superconducting) or 77°K (high-temperature superconducting). Technical
improvements and a better knowledge of dealing with and controlling cryogenic
systems have allowed SMES to penetrate the market and compete with more common
storage systems according to Sels (2001).
The dynamic performance of SMES is far superior to most other storage technologies.
Response times down to milliseconds are possible and the energy can be transferred
very quickly (limited normally by the cost of the power conversion components).
Another key feature is the virtually unlimited number of charge/discharge cycles. Sels
57
Final Report
Electricity Storage and Wind Energy Intermittency
(2001) places the overall efficiency at over 90%, while Schoenung (2001) states it can
be as high as 95%. The losses are mostly dictated by the cooling system.
SMES are most suitable for high value/low energy applications, where the storage
requirement is for less than a few seconds, with power requirements up to 1 or 2MW.
Although research is being conducted on larger SMES systems in the range of 10 to
100MW (with storage times of minutes), recent focus has been on the smaller microSMES devices in the range of 1 to 10MW for the power quality market, which are
becoming commercially available. A commercial product for example provides
approximately 1MW for 1s. Buckles (2000) reports that the discharge rate can be
easily controlled and thus the available bridging time becomes longer as the load is
smaller. The size is an obvious limitation to large systems. One estimate of the radius
of a coil supporting a load of 5,000MWh, 1,000MW is 150-500m, depending on the
peak field and the ratio of the coil height to diameter, while Breeze (1998) states that
a 5,000MW unit would need a coil of radius 800 m.
SMES technology provides an efficient protection against voltage sags (supplying
reactive power to the system) or momentary outages. They can be used for smoothing
fast changing loads at a small scale like factories or at a bigger-scale at the
distribution and transmission level according to Yurek (1999).
The projected capital cost and, to a lesser extent, the high energy consumption by the
cryogenics and refrigeration systems, make SMES unattractive for competitive
diurnal storage applications such as generation and transmission deferral, load
levelling, peak reduction and renewable applications concludes Swaminathan (1997).
In continuous mode operation, the system is constantly cycled and the parasitic losses
are proportionally less. In diurnal storage applications, however, these parasitic losses
are proportionally large, thus reducing overall system efficiency according to Akhil
(1997).
Table 3.3 shows an estimation of the costs of a 1MW unit in 2001 and the expected
evolution in 5 and 10 years from Sels (2001). The cost increases as the bridging time
becomes longer. The O&M costs are not very well established, but are expected to be
around $8/kW-y, including refrigeration according to Gordon (1995). The specific
costs of SMES facilities fall as the size increases.
discharge time
2001
2006
2011
1s
865
224
178
30 s
1388
403
344
60 s
943
540
464
Table 3.3 Estimated cost for a 1MW SMES unit (×€1,000)
The deployment of SMES demands a reduction of costs, possibly achieved by the
design of high-temperature super-conducting materials and low-temperature power
electronics. Siemens completed an evaluation and conceptual design of a 2MWh/50
MW SMES for use in providing frequency stabilisation to the electric system, but
Prescher (1995) reported that it may be too expensive compared to other storage
technologies. The PCS represents more than 60% of the total cost, but decreases
58
Electricity Storage and Wind Energy Intermittency
Final Report
significantly with increasing rated power and decreasing bridging time. Another
problem that must be addressed is the stability of the superconducting coil, which is
very sensitive to small temperature deviations (Sels, 2001).
Figure 3.5 Distributed SMES at a substation
American Superconductor is commercialising the system D-SMES, with a power
rating of 1MW, to provide grid stabilisation and support (Price, 2000). These units
can provide up to 3MW of instantaneous real power and up to 8 MVAR of reactive
power from the converters. The energy capacity is 3 MJ, which means at full power
the discharge lasts 1 s. The charge is carried out in less than 90 s. Six of these units
are being used to stabilise a 115kV transmission grid in Wisconsin, avoiding grid
upgrades.
3.6
Batteries
Batteries are the most common devices used for storing electrical energy. Battery
Energy Storage (BES) can be seen as the ‘standard’ storage system. There are a
number of different technologies with their own reactions, materials and electrical
characteristics. This wide variety of attributes leads to tremendous diversity in battery
types and uses. Some of these technologies are presently applied in electric power
applications, including utility-scale energy storage facilities. Despite the interest of
utilities in BES, end-users continue to be the largest market for batteries.
As battery cells have a characteristic operating voltage and maximum current
capability, battery systems normally consists of several cells, linked in line or parallel
dependent on the required power and energy rating.
Batteries exhibit a fast response to changes in power demand. Their efficiency varies
among technologies, and also depends on the application and the operation regime.
Applications
BES shows the broadest application range. Their response time is suitable for
virtually all applications while modular build-up ensures great flexible and
enlargement. The flexibility of batteries is illustrated in the range of applications for
59
Electricity Storage and Wind Energy Intermittency
Final Report
which they have been used, from the 10MW / 40MWh installation in Chino to the
sub-kW/kWh UPS systems available for computer applications.
BES has been receiving considerate attention from the utility industry. During the
eighties they were considered a viable option for large-scale load levelling
applications. However, there is an increasing controversy as to whether they can be
viewed as energy supply technology, serving applications such as load levelling and
capacity deferral, as the economics mitigate against operating in this manner. The cost
structure of BES makes them less competitive for applications that require high power
(MW scale) for long durations (> 1h). Large installed BES systems have typically
power-to-energy ratings resulting in a discharge durations from ½ hour to 4 hours.
Although there are systems with longer discharge times, the competitiveness against
non-storage technologies is progressively dubious. Swaminathan (1997) found that
this trend could be observed in the 90’s, with installations which have large power
ratings but are designed to operate for durations typically < 1 hour. Lower discharge
time limits are set by the discharge characteristics of the battery.
Utilities do see key roles for batteries especially in areas such as distributed
generation and power quality, but Akhil (1997b) notes that they still express their
concern about costs, life span, maintenance, and energy density. It must be stated
however, that although battery technology is mature, stationary battery systems are
not. System costs still need to be reduced, and it is anticipated that this can be
achieved through optimised integration and mass production.
BES systems are also widely used in small off-grid renewable back-up applications.
They can also be seen as an option for renewable energy integration, increasing the
reliability and dispatchability of renewable sources.
Battery energy storage systems must be optimised to give the best possible
performance for a given application. This optimisation may include the actual design
of the battery cell and so it is important to understand the different parameters
affecting battery design. These include:
• power requirements
• energy requirements
• charge/discharge rates
• number of discharge cycles
Classical vs. advanced batteries
The most mature technology, flooded lead-acid (LA) batteries and valve regulated
lead-acid (VRLA) batteries, have been in service in electric power applications for
two decades according to Butler (2002). Nickel-cadmium (NiCd) batteries have also
reached an important maturity degree.
Lead-acid batteries have a very high efficiency and all other electrical storage
technologies have to compete with them. However their low power and energy
density and other drawbacks have encouraged much research to develop new battery
technologies. The advanced battery technologies offer improved power and energy
densities, as well as lower maintenance, but Gyuk (2002) found that they do not have
such a proven track record and tend to be expensive for large-scale applications. They
60
Electricity Storage and Wind Energy Intermittency
Final Report
are being used mostly in vehicles and for power quality and backup purposes at
manufacturing plants.
Advanced battery technologies such as sodium-sulphur (NaS) and lithium-ion are
quickly becoming commercially available. Lithium-polymer (Li-polymer) and
nickel-metal hydride (NiMH), which have been developed mainly for automotive
use, and metal-air, are also candidate storage media, but are just emerging in pilot
scale systems according to Butler (2002).
Good battery management and well-optimised operational regime are important for
the financial viability. Not all batteries are the same and so care should be taken in the
application requirement specification to ensure that the requirements placed on the
battery storage media are well matched to the battery specifications (number of
anticipated discharges, depth of discharge, rate of discharge, etc.).
The costs vary a lot between technologies. Based only on the initial costs, the least
expensive is lead-acid followed by NiCd. Advanced battery technologies are currently
more expensive, but costs are expected by Collinson (2000) to fall as the volume of
sales increases. O&M costs hold an important share of the life-cycle costs in most
technologies.
Some batteries involve environmental hazards. Lead and cadmium, for instance, are
highly toxic elements.
Information about battery manufacturers and other related issues may be found at
www.basytec.de/links_e.html.
Lead-acid batteries
The majority of operating energy storage worldwide in both utility and non-utility
applications is in the form of flooded lead-acid (LA) and valve regulated lead-acid
(VRLA) battery technology. This success is founded on its maturity, relatively low
cost and long lifespan. Due to the fast response and reasonably low self-discharge
rate, they offer a very flexible solution for energy storage, ranging from short-term
applications (seconds) to long duration storage (hours). A major disadvantage is the
low performance quotient in terms of energy and power densities, which may pose a
problem in applications where space is a serious constraint. The other major drawback
is the high O&M requirements. Furthermore, there is very little cost reduction margin
for lead-acid batteries in the future. However, due to their current costs and the
familiarity of industry with them, they will always be an option for less taxing
applications.
Although it is anticipated that incremental performance improvements will continue
to be made with conventional lead-acid battery technology, Collinson (1999)
maintains that future advances in system performance will probably come from
developments in system operation, control and design optimisation, including the
energy storage systems that can provide multiple benefits.
VRLAs use the same basic electrochemical technology as flooded LA batteries, but
they are closed with a pressure-regulating valve, so that they are essentially sealed.
Therefore there is no venting of hydrogen and oxygen, and no ingress of air into the
61
Electricity Storage and Wind Energy Intermittency
Final Report
cells. In addition, the acid electrolyte is immobilized, which eliminates the need to
add water to the cells to keep the electrolyte functioning properly, or to mix the
electrolyte to prevent stratification. They can be used close to people and sensitive
equipment. The major advantages of VRLAs over flooded lead-acid cells are the
dramatic cut in the maintenance and the reduction in weight and size. The drawbacks
are their higher cost and lower lifetime. The battery subsystem may need to be
replaced more frequently than with the flooded lead-acid battery, increasing the
levelised cost of the system. VRLAs have become popular for standby power supplies
in telecommunications applications, where they are perceived as maintenance-free
and safe, and for uninterruptible power supplies in situations where special rooms
cannot be set aside for the batteries.
There is a significant body or research attempting to address the durability and
charge/discharge rate deficiencies of LA and VRLA batteries. The international
Advanced Lead-Acid Battery Consortium has developed a technique to significantly
improve storage capacity and recharge the battery in a few minutes rather than hours.
In US, a related technique appears to have extended cycle life by three or four times.
In general, lead-acid batteries, both flooded and valve-regulated, are a popular storage
choice for power quality, UPS and some spinning reserve applications. Its application
for energy management, however, has been very limited due to its short life cycle.
They have been used in a few commercial and large-scale applications. The largest
one is a 10MW /40MWh system in Chino, California, built in 1988 and shown in
figure 3.6. The LA battery system operated by PREPA in Puerto Rico since 1994,
which provides spinning reserve as well as voltage and frequency control for the
island's grid, has a larger power (20MW), but a lower energy capacity (14MWh). This
facility provides spinning reserve in case of a generator outage –its quick response
time enables the system to maintain a smaller spinning reserve capacity– and contributes
to frequency regulation for the San Juan metropolitan area. The economic and
technical success of this system has resulted in a decision to double the capacity of the
system. However, premature aging of the batteries, initially projected to last for 10
years, suggests a cycling rate that is too demanding for the batteries.
Figure 3.6 10MW/40MWh Lead-Acid storage system in Chino, California
The 1MW/1.4MWh VRLA plant in Metlakatla Island (Alaska) supports minigrid
stability. Berlin Power and Light (BEWAG) operates an 8.5MW/8.5MWh and a
62
Final Report
Electricity Storage and Wind Energy Intermittency
17.5MW/5.7MWh plant in a combined spinning reserve and frequency regulation
mode. The 3MW/4.5MWh VRLA plant at Vernon provides back-up power to critical
loads and peak shaving to a factory.
Table 3.4. lists and compares some lead-acid (including VRLA) storage systems
larger than 1MWh. Since some plants have a greater energy capacity or a greater
discharge rate, Akhil (1997) warns that it is difficult to compare the cost of batteries
on a $/kW or a $/kWh basis. The Chino battery, which is three times larger than
PREPA and 9 times larger than Vernon, appears to have benefited from economies of
scale with a cost of $20l/kWh.
Plant
Rated
Year
of
Energy
installation
(MWh)
Rated
Power
(MW)
Battery
system
alone
Cost in Cost in
$1995
$1995
($/kWh) ($/kW)
CHINO
1988
40
10
201
California
HELCO Hawaii
1993
15
10
304
(VRLA)
PREPA Puerto
1994
14
20
341
Rico
BEWAG
1986
8.5
8.5
707
Germany
VERNON Calif.
1995
4.5
3
305
(VRLA)
* Includes Power Conditioning System and Balance-of-Plant
Total cost of the
storage system*
Cost in Cost in
$1995
$1995
($/kWh) ($/kW)
805
456
1,823
456
777
1,166
239
1,574
1,102
707
n/a
n/a
458
944
1,416
Table 3.64 Largest LA and VRLA batteries installed worldwide
The total cost of a BES for large storage applications generally range from $1,200 to
$1,500/kW for a 1-2 h system, whereas the specific cost of a system for power quality
applications is around $450/kW. Akhil (1997) estimates a cost reduction potential of
around 20%. At any rate, the costs are driven by a combination of power and energy
ratings. Schoenung (2001) gives a power-based specific cost of $250/kW, and an
energy-related cost of $225/kWh, which can be as low as $100/kWh for power quality
applications.
Wagner (1999) points out that utility-scale lead-acid batteries can be also installed in
connection with renewable energy plants, such as the multifunctional 1.2MWh plant
installed in combination with a 2MW wind farm in Bocholt (Germany), which is used
for UPS, improvement of power quality and peak-load shaving.
Developers / Suppliers: GNB Industrial Power/Exide, East Penn, Trojan, Crown
Battery.
Nickel-cadmium batteries
A nickel-cadmium (NiCd) battery uses an alkaline electrolyte, usually potassium
hydroxide (KOH) or occasionally sodium hydroxide (NaOH), which act as an ion-
63
Electricity Storage and Wind Energy Intermittency
Final Report
conducting medium. NiCd batteries cost more than LA but have many advantages.
They exhibit higher specific energy and longer lifetime, require less maintenance,
endure more extreme conditions, and can withstand full discharge without
compromising the battery life and efficiency. An important concern is the toxicity of
cadmium, which causes severe problems for the disposal of the batteries. Cadmium
might even be banned in the future. Other drawbacks include high rates of selfdischarge and ‘memory’ effect, which reduces the storage capacity available.
A 40MW NiCd plant comprising 13,760 cells from SAFT is under construction for
voltage support on a long power line to Fairbanks, Alaska. Although these batteries
are not common for large stationary applications, resistance to cold may have been
among the deciding factors in this case. The system, completed in September 2003, is
able to supply 46MW for 5 minutes, but will typically provide 27MW for 15 minutes.
The cost of the facility was $35M.
Lithium-ion batteries
The main advantages of Li-ion batteries, compared to other advanced batteries, are
their high energy density (four times that of lead-acid batteries), very high efficiency
(near 100%), and long life cycle (3,000 cycles at 80% depth of discharge).
High energy density enabled Li-ion to take over 50% of small portable market in a
few years, but there are some challenges in upscaling to large-scale batteries. The
main hurdle is the high cost (above $600/kWh) due to special packaging and internal
overcharge protection circuits. Several companies are working to reduce the
manufacturing cost of Li-ion batteries to capture larger energy markets, mainly the
auto industry but also multi-kW, kWh sizes for residential, commercial, and
renewable markets. However, only power quality applications and short-duration peak
shaving seem to be feasible in the medium term according to Boyes (2000).
Major Li-Ion battery manufacturers include SAFT, Sanyo Electric Company, and
Hithachi.
Sodium-sulphur batteries
A special case among advanced batteries is the NaS battery. Developed in Japan, this
battery operates at high temperatures. The electrodes are liquid –molten sodium for
the cathode and molten sulphur for the anode– and the electrolyte solid –alumina
ceramic. The assembly has to be maintained at 300ºC to keep the electrodes molten.
Extensive tests have demonstrated safe containment under extreme conditions. They
were first developed for automotive applications and are now being successfully
applied to large commercial buildings and utility power stations to provide power
quality and load levelling functions. Some utilities have also begun investigating this
technology for deferring substation upgrades.
This type of batteries is characterized by long cycle life, good energy efficiency (up to
86%), high specific energy (3–4 times a lead-acid battery), and very low selfdischarge rates. NaS batteries have a pulse power capability over six times their
continuous rating (for 30 seconds). This attribute enables the NaS battery to be
64
Electricity Storage and Wind Energy Intermittency
Final Report
economically used in combined power quality and peak shaving applications. The
continuing high cost is a major barrier for a greater market penetration.
The high operating temperature necessary for NaS batteries makes them more
efficient when serving applications in which the battery cycles frequently and the
thermal effects of cycling contribute to the maintenance of the operating temperature
of the unit. They could also serve applications with long periods of inactivity if the
same battery system also served a frequent cycling application reports Butler (2002).
Some 38 systems totalling approximately 20MW and 124MWh have been installed in
Japan. The largest of these installations is a 6MW, 8h unit for Tokyo Electric Power
company shown in figure 3.7, which is used for long-term electricity storage (loadlevelling and load management). It also can supply active and reactive power to
mitigate voltage sags and frequency fluctuations. The peak power is up to six times
the continuous rating. The batteries of this plant alone exhibit efficiencies of 86%
according to the developers. The entire system, including the AC/DC converters, has
an efficiency of about 75%. A second system of the same characteristics was
commissioned subsequently. Outside of Japan, operation of a 500kW demonstration
unit has been installed in the US to be used for load levelling or as an UPS according
to Gyuk (2002).
Figure 3.7 6MW, 8h NaS battery storage system
Prospects for this technology are focused on the retail market for energy management
and power quality. Combined power quality and peak shaving applications is an
important potential market. Commercial production of the basic building block –the
NaS 50kW, 360 kWh module– is the initial target.
Developers / Suppliers: NGK
Metal-air
Potentially the cheapest batteries are metal-air batteries, which, along with their very
high energy density, explains why a good number of companies including Evonyx,
AER Energy Resources, Metallic Power, Chem Tek, Power Zinc, Electric Fuel,
Alupower and Aluminium Power are developing them. Anodes contain commonly
available metals, like zinc, aluminium, lead and even iron, placed in a liquid or
65
Final Report
Electricity Storage and Wind Energy Intermittency
polymer impregnated electrolyte of potassium or other conducting hydroxide. The
cathodic material is oxygen from ambient air. Metal-air batteries are inherently safe
and environmentally benign. But while high energy, controllable discharge and low
cost could suit them to many primary battery applications, the only known
rechargeable unit so far available, a zinc-air system, has a very short cycle life and a
charge/discharge efficiency of only about 50%. Thereby the technology demands
further research and development before it can compete with other types of batteries.
Zinc-air batteries are being developed by Evonyx.
Comparison table
Table 3.5 summarises some relevant data for battery technologies comparison.
Battery
type
Lead-acid
NiCd
NiMH
Li-ion
Li-polymer
NaS
Zn-air
Energy
density
(Wh/kg)
10-20
30-37
75
150
200
53-116
120-180
Energy
density
(Wh/litre)
50-70
58-96
240
400
220
40-170
169-180
Operating
temp (°C)
-10 to 40
-40 to 50
-20 to 50
-20 to 50
310 to 350
Efficiency
(%)
85
65
65
95
65
75-86
50
Selfdischarge
Cycle
(%
loss life
/month)
(cycles)
variable
10
< 2000
15-25
up to 600
2
3000+
1000+
0
2250+
200
Table 3.5. (Source: Linden, 2001)
3.7
Flow batteries
Flow Batteries (FB), also known as Regenerative Fuel Cells or Redox Flow Systems
are a new class of battery that has made substantial progress technically and
commercially in the last years. Flow Batteries Energy Storage (FBES) systems have
features that make them especially attractive for utility-scale applications. The
operational principle differs from classical batteries. The latter store energy both in
the electrolyte and the electrodes, so to speak. Flow batteries, however, store and
release energy using a reversible reaction between two electrolyte solutions separated
by an ion permeable membrane. Both electrolytes are stored separately in bulk storage
tanks, whose size defines the energy capacity of the storage system. The power rating
is determined by the cell stack. Therefore the power and energy rating are decoupled,
which gives the system designer an extra degree of freedom when designing the
system. The cost per kWh decreases as the energy storage capacity increases for a
given power, and can reach very competitive values. Since the energy storage
capacity depends exclusively on the size of the electrolytic tanks, the flow batteries do
not have obvious scale limits. This makes them a promising candidate to join CAES
and PHES providing large-scale energy storage.
FBs are flexible in operation, especially with respect to discharge times, which can
range from minutes to many hours. Other interesting features are the fast response
delivering real power (reactive power is not delivered so quickly) and the capability
of withstanding overload and total discharge without any risk of damage (Sels, 2001).
66
Electricity Storage and Wind Energy Intermittency
Final Report
Flow batteries are generally not more than between 75% and 80% efficient. This is
higher than electrolysis-fuel cell systems, but still below the efficiency of most
batteries. The need for pumps to circulate the electrolytes adds a parasitic loss, but at
the same time makes the thermal management easier (heat exchanger in the
circulation loop). Another disadvantage is the use of aggressive chemical solutions.
FBs are entering the market now, and are likely to experience sustainable growth to
attain commercial-scale production in 3-5 years. Developers are positioned for mass
manufacturing, are soliciting large-scale orders, and report significant interest among
potential buyers (Lotspeich, 2002). Demonstration FBs have performed well in a
range of applications that showcase their technical versatility and the potential
economic benefits of providing multiple services and value streams. FBs are an
integrative technology, serving a range of fragmented storage market niches such as
load levelling, peak shaving, power reliability, and substation-based transmission
support.
Many different electrolyte couples have been proposed for use in flow batteries.
Present developments are based on:
• Vanadium Redox
• Sodium polysulphide / Sodium bromide
• Zinc/Bromine
Vanadium Redox Flow Battery
Vanadium Redox Batteries (VRB) store energy by employing vanadium redox
couples (V2+/V3+ in the negative and V4+/V5+ in the positive half-cells). These are
stored in mild sulphuric acid solutions (electrolytes). During the charge/discharge
cycles, H+ ions are exchanged between the two electrolyte tanks through the
hydrogen-ion permeable polymer membrane. The net efficiency of this battery can be
as high as 85%. Efficiencies over 80% have been proven, according to Blackaby
(2002). Like other flow batteries, the power and energy ratings of VRB are
independent of each other, which gives the flexibility to increase the system capacity
by simply increasing the volume of solution. Menictas (1994) affirms they can be
fully discharged (100%) without any detrimental effects. The VRB offers a very long
lifetime, fast response (from charge to discharge modes in 1/1000 s) and high
overload capacity (more than twice the rated power for several minutes). The
technology has proven its capability for ride-through, power quality and emergency
back-up.
Cells last at least 10,000 cycles, and in the laboratory 25 kW modules have exceeded
16,000 cycles according to Lotspeich (2002). The electrodes are made of inert
materials. Stack service life is determined primarily by membrane longevity, and the
life of pumps and other auxiliary components. The manufacturer Sumitomo Electric
Industries (SEI) recommends that the stack be replaced be replaced every 10 years,
reflecting an expected membrane life of 8–10 years. The electrolytes have an
indefinite life, and may be reused.
67
Electricity Storage and Wind Energy Intermittency
Final Report
Figure 3.8 VRB at a semiconductor factory
VRB was pioneered in the Australian University of New South Wales (UNSW) in
early 1980's. The Australian Pinnacle VRB bought the basic patents in 1998 and
licensed them to SEI and Vanteck. Another developer is Cellenium Co Ltd. To date,
the majority of the demonstrations have been on-grid load-levelling and peak-shaving
applications in Japan, where storages up to 500kW, 10 hrs (5MWh) have been
installed by SEI. VRBs are used for load levelling at a substation (450 kW, 900 kWh),
a university (500 kW, 5MWh), and an office building (100 kW, 800 kWh), and
stabilise the output of both photovoltaic (30 kW, 240 kWh) and wind generators (170
kW, 1.2MWh).
This latter plant, provided by SEI, can supply 170 kW for 6 hour and is installed with
a 275 kW wind turbine. VRBs have also been applied for power quality applications –
figure 3.8 shows a VRB facility installed by SEI at a Japanese semiconductor factory
provides 1.5MW for load levelling, and can yield 3MW for 1.5 sec to eliminate
sudden sags. The first large commercial VRB outside Japan was installed in South
Africa by Vanteck (250kW, 520 kWh, ~2 hrs) for UPS ride-through, power quality
and emergency back-up. PacifiCorp built recently a 2,000 kWh (8 h) system in a
remote area in Utah, the first in North America, to provide peak power and end-ofline voltage support, deferring the need for a new substation. VRBs are scalable to
MW sizes, and Lotspeich (2002) points to studies on feasible systems up to 100MW.
Developers / Suppliers: Pinnacle, Sumitomo Electric Industries, Vanteck
Technology Corp & Telepower Australia, Cellenium Company Limited.
Polysulphide bromide flow battery
Polysulphide Bromide Batteries (PSB), also known as Sodium/Bromide, are a
regenerative fuel cell technology based on a reversible electrochemical reaction
between two salt solution electrolytes – sodium bromide and sodium polysulphide.
PSB electrolytes are brought close together in the battery cells where they are
separated by a polymer membrane that only allows positive sodium ions to pass
through. Cells are electrically connected in series and parallel to obtain the desired
levels of voltage and current. A 100 kW module for instance comprises a stack of 200
cells, and sixteen m3 of each electrolyte is needed for each MWh of storage, according
to Lotspeich (2002).
Lotspeich (2002) states that the net efficiency ranges from 55–60% to about 75%
depending on its operational mode, including power conversion and energy losses due
to auxiliary equipment such as pumps. It operates ideally between 20–40°C, but
68
Electricity Storage and Wind Energy Intermittency
Final Report
tolerates a wider temperature range. The energy density is 20 to 30 Wh per litre,
which is about 30% of a normal lead-acid battery. Therefore they are relatively large
and heavy.
This flow battery was developed during the 1980s–90s by the British firm Regenesys
Technologies Ltd., owned by Innogy, in turn owned by its German parent RWE. The
patented name of the flow battery is Regenesys®.
Figure 3.79 Scheme of a Polysulphide Bromide Battery
The power conditioning system and controls linking Regenesys® to the grid allow
‘cold’ start up in less than 10 minutes. Lotspeich (2002) reports that if held in standby
mode with charged electrolyte in the stacks, the system can respond in fraction if a
second (reportedly within 20 milliseconds) to supply more than 10MW. It is designed
to be automated and run remotely, with biweekly removal of sodium sulphate crystal
by-products.
In the year 2000, Regenesys Technologies began building a 15MW, 120MWh energy
storage plant (the largest battery in the world) at Innogy's Little Barford Power Station
in the UK, following the success of a four-year, 1MW demonstration at Innogy's
Aberthaw power station. It was expected to come into operation in 2003 and to last
for at least 15 years with overall cycle efficiency predicted at 60-65%. The purpose
was to provide a combined-cycle gas turbine power station of the same power with
black start in the event of a grid supply outage. Other duties of the plant included
frequency response and voltage control, as well as energy management in the UK’s
electricity market.
The use a RFB in parallel with a power plant was expected to increase the overall
efficiency by avoiding inefficient partial loads. At the same time, expensive start-up
and shutdowns of other plants could be delayed. It was also hoped to use the plant at
Little Barford in conjunction with wind power generation according to Price (2000b).
The facility cost was approximates at €21-million, that is €1400/kW or €175/kWh.
The system was anticipated to reach full charge or discharge power in 0.1s, asserts
Price (2000b). This, for a first commercial scale plant, would be very encouraging, as
costs should fall significantly with volume and further development. The O&M costs
were estimated to be relatively small, according to Semadeni (2003). The facility,
69
Electricity Storage and Wind Energy Intermittency
Final Report
which includes two tanks of 1800 m3, covers an area of nearly 5000 m2 as shown in
figure 3.10).
Figure 3.10 Drawing of the 120MWh planned Regenesys plant at Little Barford
Tennessee Valley Authority (TVA) ordered a 12MW, 120MWh Regenesys® unit
(USA) to be operational in Mississippi in late 2004 (shown in figure 3.11), at a total
cost of $25 million. The expected operational life of the plant was 15 years and the
purpose was to reduce the need to build additional power lines and power plant for
peak generation. TVA also felt it would enhance the reliability and improve power
quality by eliminating momentary interruptions in service and maintaining voltage
levels. TVA assessed the economic viability of a storage system against the
installation of low capacity factor peak generation, as well as grid upgrades to prevent
interruptions (TVA, 2001). However, in 2002 TVA cancelled plans for a second
Regenesys plant, which was to be associated with a 20MW wind farm.
Figure 3.11 Electrolyte tanks of the TVA Regenesys plant under construction in
Mississippi
In response to a query from the report authors in 2003, the Regenesys Technologies
Marketing Manager confirmed there were some international orders in various stages
of negotiation. Lotspeich (2002) states that systems of up to 500MW capacity are
feasible in its current configuration. Regenesys Technologies were targeting the
renewable generation market, with the conviction that the system can be economically
feasible in enabling better use to be made of intermittent sources, availing of the
differential between peak and off-peak rates. Innogy planned MW-scale support for
70
Electricity Storage and Wind Energy Intermittency
Final Report
wind farms in Denmark and elsewhere, and reports customer interest in installations
ranging from 12–100MW according to Lotspeich (2002).
Despite the positive signals, Innogy abandoned its Regenesys electricity storage
project in December 2003 according to The Guardian Newspaper, after its German
parent, RWE decided against investing the money need to commercialise the plant.
This system seemed very suited to addressing the intermittent nature of wind energy
in Ireland and the reasons for the abrupt cessation in development are currently
unclear.
Developers / Suppliers: Regenesys Technologies Limited (ceased in December
2003)
Zinc bromine flow battery
ZnBr batteries are also FBs, but their design and electrochemistry differ from the
PSBs and VRBs. A Zn-Br electrolyte flows through two half-cells divided by a
microporous membrane, with a Zn- electrode and a Br+ electrode. Unlike other FBs
and to an extent similar to conventional batteries, the electrodes serve as substrates for
the reactions and their performance capacity can be degraded if the battery is not
completely and regularly discharged according to Lotspeich (2002).
During charge, zinc is electroplated on the anode and bromine is evolved at the
cathode. An agent in the electrolyte is used to reduce the reactivity of the elemental
bromine by forming a polybromide complex, thus minimising the self-discharge of
the battery. The complexed bromine is then removed from the stacks via the flowing
electrolyte and is stored in the external reservoir. On discharge, the complexed
bromine is returned to the battery stacks and reduced to bromide on the cathodes,
while zinc is oxidized to zinc ions on the anodes.
The battery has a lifetime of up to 2000 cycles, and can be repeatedly deeply
discharged (100%) without noticeable performance deterioration. The system’s net
efficiency is about 75%.
The ZnBr battery was developed by Exxon in the early 1970's. Over the years, many
multi-kWh ZnBr batteries have been built and tested. A 1MW/4MWh system was
implemented at Kyushu Electric Power in 1991. The primary developer is the
U.S./Australian firm ZBB Energy. ZBB combines 25 kW, 50 kWh standard modules
with PCS and controls into containerised 250 kW, 500 kWh turnkey units; Lotspeich
(2002) states that power output can be doubled with appropriate PCS. These systems
are suitable for industrial energy storage, or may be combined into larger multimegawatt sizes for utility applications. Zinc-bromine batteries are thus available off
the shelf, although Gyuk (2002) points out that integrated power electronics are
essential to successful applications. U.S./Austrian developer Powercell sold 100 kW,
100 kWh modules, but shut down in April 2002.
Installations have been made in several countries, providing facility-scale UPS and
load management and supporting microturbines and solar generators as well as
substations and T&D grids. The plants installed include several remote villages in
Malaysia, powered by hybrid off-grid systems. Systems have been demonstrated on
71
Electricity Storage and Wind Energy Intermittency
Final Report
trailer-mounted mobile systems at both 1.8MW, 1.8MWh and 200kW, 400kWh
scales. ZBB’s baseline turnkey product is a 500kWh system. In December 2003, ZBB
secured a contract to supply four 500kWh systems to Pacific Gas and Electric
Company.
Figure 3.12 Containerised 200 kW, 2.5 h ZnBr battery
The technology is modular, as shown in figure 3.12. Modules can be linked
electrically but not hydraulically. Hermetic electrolyte tanks isolated within each
module limit economies of scale in larger installations of aggregated modules. ZBB
sizes PCS and control systems to serve the number of modules in each application.
ZnBr batteries are viewed as a candidate for renewables backup. Transportability, low
weight, and flexible operation are advantageous when compared to lead-acid batteries,
and can outweigh the somewhat lower electrical efficiency.
ZBB is targeting the renewable integration market with a module especially designed
to operate in connection with renewable sources. ZnBr batteries can be used to
smooth out the wind farm fluctuations and hence help to control the frequency
fluctuations. The company is in negotiations with Apollo Energy Corporation to
provide 30 ZnBr batteries to back up a 20MW wind farm for several minutes. The
goal is to keep the wind farm operational for the 200+ hours each year when erratic
winds would otherwise force operators to shut down some turbines.
Developers / Suppliers: ZBB Energy Corp
3.8
Hydrogen energy storage
Hydrogen is envisaged as a promising means of electrochemical storage. In a
Hydrogen Energy Storage (HES) system, the charge takes place when the electrical
energy is used in an electrolyser to split water into hydrogen and oxygen. Although
oxygen has an economic interest, it is usually vented to the atmosphere. Hydrogen can
be stored in different ways. The discharge, providing the energy release, can take
place in a fuel cell or in an internal combustion engine (ICE).
One advantage of hydrogen storage systems, compared with the others discussed, is
that the energy storage capacity and input/output power rating are completely
decoupled. For many electrical energy storage applications, this advantage is
significant. This notwithstanding, it is the expected use of hydrogen in the future as a
72
Electricity Storage and Wind Energy Intermittency
Final Report
fuel for transport and other applications that provides the greatest appeal for hydrogen
as a means of storage.
Hydrogen energy storage systems have many drawbacks. Most aspects in the
hydrogen-related technology, including generation, storage and utilisation in fuel
cells, need further development. HES appear as an option for the long run on account
of the current modest scale and relative immaturity of the components, and the
consequent present high costs. A lot of research is still necessary to make storage
through hydrogen attractive for utility-scale storage.
The most significant challenge facing HES is the low round-trip efficiency. There are
losses in the electrolyser, in storage and in the fuel cell. Technological breakthroughs
will improve the efficiency, but it will still remain considerably behind other
competing technologies. Future estimates of electrolyser and fuel cell efficiencies
vary quite a lot among the different literature sources. The hydrogen storage option
chosen will also influence the efficiency. It is therefore not easy to provide accurate
estimates of the efficiency and its likely evolution. This notwithstanding, a reasonable
efficiency range is of the order of 34-40% for FC-based systems and 29-33% for those
using ICEs for discharge.
Swaminathan (1997) reports on a study carried out in 1997 by the US Department of
Energy, in which hydrogen was identified as a candidate for long duration storage
applications such as load management, peak shaving and transmission & distribution
capacity deferral. However, HES was compared to batteries both in a current and
future scenario, and despite the optimistic projections used for the hydrogen storage
component costs and the more than optimistic round trip efficiency of 50% projected
in the long term, HES was only find to break even with batteries for applications
requiring 5.4 hours or more.
Due to the modularity and availability of small-scale components, and the decoupling
of power and energy capacity, HES is particularly attractive for long-term storage in
renewable-based stand-alone systems. This may in fact be the first market that HES
will penetrate.
Despite the concerns that exist regarding safety, hydrogen does not pose more of a
problem than other fuels. Being the lightest gas, hydrogen quickly disperses into the
environment in the event of leakage, making it a lower fire hazard than gasoline.
Extensive research has been carried out to simulate consequences of hydrogen
releases in confined and unconfined spaces. So far this has not revealed any issues
that are considered to be unacceptable or which represent unmanageable risk.
Components of a HES
Electrolyser
Generation of hydrogen via electrolysis is a well-known and established technology.
It is mostly used when moderate amounts of high-purity hydrogen are required. Due
to the high cost of electrical energy, only a very small proportion of the worldwide
hydrogen production comes from electrolysis, but the foreseen changeover to the
73
Electricity Storage and Wind Energy Intermittency
Final Report
hydrogen economy indicates an ever-growing demand of electrolysis-generated
hydrogen, especially using renewable power.
Electrolysers serving in electrical storage applications will operate with fluctuating
input power, and must be designed to meet the following requirements:
• high efficiency
• good dynamic performance
• possibility of operating over a wide input power range with high current yields
and sufficient gas purities
• durability under frequently changing conditions of operation
These features must be attained at the lowest cost possible. Electrolysis technology in
general has experienced some advances during the last years, achieving higher
efficiencies –up to 85% based on the Higher Heating Value (HHV) of hydrogen– and
longer-lasting stacks. Cognisant of the great potential market for renewable
electrolysis, manufacturers have managed to enhance the performance of their
electrolysers under conditions of variable power input.
Electrolysers are inherently modular devices, since the production capacity is
proportional to the number of cells that make up the stack. The specific cost declines
as the size increases. The largest commercial systems can produce 485 Nm3/h, which
corresponds roughly to an input power of 2.5MW. Larger customer-tailored
electrolysers can also be produced, although the cost rises.
Electrolysis has been traditionally based on an alkaline technology, but Proton
Exchange Membrane (PEM) stacks are now coming to the forefront. Figure. 3.13
shows an example of a PEM electrolyser and storage vessel. Higher hydrogen purity,
faster dynamic response, lower maintenance and increased suitability for
pressurisation are among the advantages of PEM technology. Efficiencies are not
currently very high however, and only small units are commercially available.
Figure 3.813 PEM electrolyser and storage vessel
Since the most likely storage solution, at least in the medium term, is pressurised gas,
another challenge to perform the electrolysis at a high pressure. If pressure can be
provided in the electrolyser and thus hydrogen delivered at high pressure, expensive
and unreliable compressors may be omitted. Commercial units rarely surpass some
74
Electricity Storage and Wind Energy Intermittency
Final Report
tens of bars (~30 bar), whereas some prototypes have reached up to 120 bar (Meurer,
1999).
The cost of electrolysers varies significantly among different manufacturers and
different scales. Divergent estimates can be found in technical literature. A specific
cost of 1,100 €/kW may be a rather optimistic estimate at present, especially in the
low power range, but realistic for large-scale systems (several megawatts) in the
medium term. In fact, specific costs might drop sharply in the future down to 600
€/kW or even 300 €/kW according to the most optimistic predictions (Krom, 1998).
Annual maintenance costs are approximately 3% of the capital cost. Lifetime is
difficult to predict on account of the limited field experience of electrolysers operating
under fluctuant conditions. The stack may last from 5 to 10 years, while the rest of the
system has a longer durability (20 years).
Developers / Suppliers: Nosrk Hydro Electrolysers, GHW, Hydrogen Systems,
Stuart Energy, Proton Energy, Casale Chemicals.
Power generation: fuel cells
There are two competing technologies for power generation from hydrogen: internal
combustion engines (ICE) and fuel cells (FC). ICE appears as a transition
technology while fuel cells are improved and costs are brought down. The
modifications that must be introduced in gas engines to be adapted to hydrogen are
not very significant.
FCs are more efficient and reliable than ICEs, and need less maintenance (no moving
parts). While ICEs give rise to some NOx emissions, fuel cells are virtually emissionfree when fuelled directly with hydrogen. Expected life spans range from 15-20 years,
although current realisations show much more premature aging of the electrochemical
cells, which can be replaced independently. It is anticipated that costs will become
competitive, when economies of scale are achieved. Fuel cells are expected to play a
major role in future energy supply. The current level activity in FC research and
development is significant, mainly driven by the transport sector but also for
stationary applications.
Fuel cells have efficiencies at partial loads higher than at rated power. This makes
FCs very attractive and efficient for applications with highly variable loads. One of
the characteristics of FC systems is that their high efficiency is little affected by size.
This, together with their modularity, low emissions and low noise level, makes them
very suitable for distributed generation. As a result, initial stationary plant
development has been focused on several hundred kW to low MW capacity plants.
It is clear that all FC costs at present – and these are estimated at anything between
500 and 8,000 €/kW are high because they are representative of an emerging
technology.
There are a number of different FC technologies, differing in the type of electrolyte.
These are described in more detail in Gonzalez and Ó Gallachóir (2003). High
temperature FCs have a significant thermal inertia. Since fast dynamic behaviour is an
important requirement in electrical energy storage applications, low temperature fuel
75
Electricity Storage and Wind Energy Intermittency
Final Report
cells are more suitable. This market segment is likely to be dominated by Proton
Exchange Membrane Fuel Cells (PEMFC), as shown in figure 3.14, which operate
at temperatures <90ºC.
Figure 3.814 1MW power plant in Alaska (5 PEMFC 200 kW units in parallel)
PEMFC is the technology being focussed on in the transport sector, not only because
of the dynamic response and rapid start-up but also the high current densities and low
weight. This will hasten as well the deployment of hydrogen-fuelled PEMFCs for
stationary applications, in which there are less stringent requirements on overall cost,
space and hydrogen storage.
Systems of up to 300 kW are currently under development and have started to be
commercialised. Costs are high, in part due to the high catalyst loading (Pt in most
cases) required for both the anode and cathode, but the high manufacturing volumes
required for transport applications, together with technological breakthroughs, is
expected to result in lower capital costs in the long term. Efficiencies are in the 4045% range when directly supplied with hydrogen.
Among the high-temperature fuel cells, Solid Oxide Fuel Cell (SOFC) are making
great advances. Current systems fuelled with natural gas achieve efficiencies from
50% to 60%, clearly superior to PEMFC. The operation temperature ranges between
600ºC and 1000ºC, which places severe constraints on materials selection, resulting in
difficult fabrication processes, and gives rise to sealing difficulties. SOFCs are very
promising, however, on the crucial issue of cost. SOFC components are potentially
cheaper and easier to manufacture that their PEM counterparts, and notable
achievements have already been made in reducing costs according to Cropper (2001).
The dynamic response of SOFC is evidently far slower than PEMFC however, which
limits to a large extent the range of applications where SOFCs can be employed.
Depending on the location of the fuel cell systems, the waste heat produced by the
fuel cell can be harnessed to improve overall efficiency and make them more cost
effective.
Developers / Suppliers: Although there is a host of smaller players in the market,
PEMFC major developers include Ballard, UTC Fuel Cells, Plug Power and Nuvera,
In the SOFC sector Siemens is developing units of several hundred kilowatts.
76
Electricity Storage and Wind Energy Intermittency
Final Report
Reversible fuel cells
Since electrolysers and fuel cells are very similar devices operating in opposite ways,
it is technically possible to combine them in a single unit. The advantage of the socalled reversible fuel cell is obviously lower capital cost. However, this must be
compared against the drawback of lower efficiency, increased corrosion and other
technical hurdles. This technology, mostly associated with the PEM concept, is still in
an early stage of development, but without any doubt is the very attractive for the
future of electrical storage systems.
Storage and compression
There are several hydrogen storage technologies, at different stages of development.
Research in this field is primarily driven by the transport sector. Compressed gas
storage is a relatively simple technology, which has the major disadvantage of low
energy density and the problems associated with mechanical compression. Liquefied
hydrogen storage is currently used for distribution, has a track record on safety, and
has moderately good volume and weight storage densities, as shown in table 3.15.
Issues regarding this technology include losses (up to 2%/day) and high energy costs
of liquefaction. Metal hydrides exhibit high volume density and are being developed
for on-board storage in vehicles, but is still an immature and expensive option.
Carbon-based absorption can achieve higher volume densities, but the costs are
even higher and there is a lack of system demonstrations. Other innovative solutions
are being investigated. Millennium Cell has developed a system in which hydrogen is
bound in a liquid chemical compound (NaBO2 + 4 H2 → NaBH4) easy to store and
transport, which would release hydrogen again when required reversing the reaction.
A different approach binds hydrogen with liquid organic hydrocarbons (Scherer,
1999).
Pressurised H2
Liquid H2
Metal hydride
Volume ratio [kWh/m3]
80 – 1000
2000 – 2800
2000 – 2400
Table 3.6 Energy density of hydrogen storage technologies
The storage efficiency is highly dependant on the choice of storage technology. It
must be borne in mind that metal-hydride and carbon-based absorption do not use
mechanical but thermal energy, releasing and absorbing it with certain losses. The
thermal energy needed could be drawn from the waste heat of other processes.
For electrical energy storage, gas pressurisation is the simplest and least costly
solution provided that enough space available. Hydrogen can be stored in pressure
vessels, or, in large-scale systems requiring long-duration storage, underground
reservoirs.
Pressure tanks are used already in the gas business, are available in a wide range of
sizes and are a well-established technology. Weight is not a concern, thus aluminium
containers can be ruled out, due to their higher cost. Therefore the vessels are made of
steel. The size of these tanks will depend on the storage requirements and the
77
Final Report
Electricity Storage and Wind Energy Intermittency
pressure. Higher storage pressures result in smaller tanks, but in higher operating
costs. Figure 3.15 shows the energy density of hydrogen as a function of the storage
pressure.
Energy density (kWh/m3)
1400
1200
1000
800
600
400
200
0
0
100
200
300
400
Storage pressure (bar)
Figure 3.15. Compressed hydrogen energy density
Schoenung (2001) estimates energy-related costs of a storage system, that is, the cost
of the pressure vessels, at $15/kWh. Other authors use a lower estimate of €11/kWh,
notably Meurer (2000). For underground storage costs in large systems, cost drop
dramatically as low as $2/kWh according to Padro (1999).
Reciprocating compressors are most commonly used for hydrogen applications.
Large double-action units have efficiencies in the 65%-70% range. Hydrogen
compressors can be electricity-powered of air-driven. The latter are safer, but a
compressed air facility is needed. The former are not readily available in the market
and are expensive.
Compression is traditionally one of the weak points of hydrogen systems
(unreliability, maintenance, lifetime of the components…). For this reason, the future
points to pressurised electrolysis, delivering hydrogen at high pressure and thus
avoiding the need for further compression. This also leads to energy savings. Figure
3.16 shows how the energy consumption of hydrogen compression varies depending
on the pressure ratio.
8%
% Hydrogen energy
7%
6%
5%
4%
3%
2%
1%
0%
0
5
10
15
20
Pressure ratio
Figure 3.816. Energy required for hydrogen compression
78
Electricity Storage and Wind Energy Intermittency
Final Report
Power electronics
Currently electrolysers and fuel cells have their own power conversion systems, with
a rectifier in the electrolyser and the fuel cell having an inverter. In an integrated
hydrogen storage system the two PCS can easily be integrated into one optimised
system. This would be particularly true for a reversible fuel cell.
Synergy: The hydrogen economy
Hydrogen is anticipated to play a prominent role as a fuel in the future, and could
eventually substitute fossil fuels completely. This is what is termed 'the hydrogen
economy'. The reasons for this fuel changeover relate to environmental concerns,
depletion of fossil fuels, supply security, etc.
The most important property of hydrogen is that it is the ‘cleanest fuel’. Emissions are
virtually zero if fuel cells are used for energy conversion. As a solution for transport,
batteries have not been demonstrated as an ideal option. Hydrogen is the most
common element in the universe, but does not exist in nature as an unbound
compound, and therefore is a secondary energy carrier –not a source– that must be
derived from other energy sources. Although in the transition to a hydrogen-based
economy, fossil fuels will constitute a major source for hydrogen production, the full
benefits of hydrogen as a clean and sustainable energy supply will only be realised
when produced from renewable energies. Water electrolysis using solar and wind
power is a sustainable hydrogen source.
Hydrogen can replace fossil fuels basically in all their applications. Hydrogen use is
expected to increase rapidly for transport and electricity generation applications, as
industry commercialises advanced technologies such as fuel cells and as costs decline.
Public and private investments on hydrogen research and development have grown
sharply during recent years. There are very ambitious hydrogen programs in the EU
(lead by Germany) USA and Japan, with many other countries showing an increasing
interest in hydrogen. Iceland is on the way to become the first 'hydrogen country' in
the world, producing hydrogen from surplus renewable energy and progressively
converting the transport infrastructure from fossil fuels to hydrogen.
Car manufacturers, electric utilities, oil companies and firms specialised in hydrogen
technology began the race to commercialise the technology a number of years ago.
All major car manufacturers have programs in place for the development of hydrogen
vehicles and infrastructure. Although the hydrogen economy is driven in first instance
by the transport sector, stationary applications will be an essential part of it.
79
Electricity Storage and Wind Energy Intermittency
Final Report
Figure 3.17. Refuelling a hydrogen car
With all this in consideration, the attractiveness of hydrogen as a storage option lies
on its multi-functionality. Stored hydrogen can be either converted back into
electricity or used as a 'zero emissions' fuel for other applications, such as transport.
Since the final goal of the hydrogen economy requires its production from renewable
sources, the HES market will be chiefly linked to renewable energy applications. In
this way, hydrogen will extend the scope of renewable energies to the transport field.
As the production of hydrogen from renewables requires the existence of a market for
hydrogen and the appropriate infrastructure, this is an option envisaged for the longterm. The need to build a new energy and fuel infrastructure is seen as a main task for
the future. The route towards a hydrogen-based economy remains unclear and the
necessary investments are huge. The infrastructure required includes production,
distribution and fuelling stations. The automotive industry and energy companies are
engaged in setting up a strategy for the progressive introduction of hydrogen into the
transport sector. Currently, the step from single prototypes to fleet demonstration
activities is being accomplished. These activities are accompanied by industrial
efforts in the field of regulations and standardisations.
The replacement of fossil fuels with hydrogen will be therefore gradual. If the
hydrogen consumed in the first hydrogen fleets comes from renewable energies, the
public perception of the hydrogen economy is likely to be more enthusiastic than if
produced from natural gas. It is envisaged that the changeover to a hydrogen economy
will not be fully realised in less than fifty years from now, but important amounts of
hydrogen will start to be used far sooner.
Projects
So far, the focus of HES has been renewable-based stand-alone systems and the
production of hydrogen from renewables. A number of demonstration projects have
been carried out [as described in Dutton (1996); Barthels (1996); Galli (1997);
Szyszka (1998); Abaoud (1998); Friedland (1999); Agbossou (2000)], in most cases
using solar energy rather than wind as the intermittent source. Many of these plants
were in connection with the International Energy Agency Hydrogen Implementing
Agreement, and more precisely, the Task XI – Integrated Systems [see Schucan
(2000)]. The scale of these test systems is modest, a few tens of kilowatts at most.
80
Electricity Storage and Wind Energy Intermittency
Final Report
In order to prove the technical viability of wind-hydrogen systems, it is necessary to
undertake demonstration projects coupling wind turbines and electrolysers in a larger
scale. Norsk Hydro Electrolysers is leading a project to provide the Utsira Island
(Norway) with a wind-hydrogen system. Also in Norway, the utility Statkraft plans to
connect an electrolysis unit to a large wind turbine. P&T Technologies and Siemens
are completing the installation in Germany of the first plant following the
development of a wind-hydrogen system patent, which includes an ICE as the
regeneration device instead of a fuel cell. The recently formed Wind Hydrogen
Limited (UK) is seeking the development of large-scale wind-hydrogen schemes and
has now two major projects under development in UK. HyGen, which is a company
formed in 1996 by firms that participated in the Clean Air Now! solar-hydrogen
project in California, is undertaking a pilot multi-megawatt commercial renewable
hydrogen generating facility and hydrogen distribution network. The facility will be
using PV as well as wind for the generation of electricity. There are also ambitious
schemes in Alaska, like a stand-alone wind-hydrogen system including an electrolysis
unit of 160 kW [see Rambach (1999)].
3.9
Power conditioning subsystem and balance-of-plant
Apart from the storage subsystem, an electrical energy storage system comprises a
power conditioning subsystem (PCS) and the balance-of-plant (BOP). It is crucial for
the viability of storage to design the power converter interface between the AC power
bus and the energy storage to be efficient, reliable and robust. A poor PCS design will
constrain the performance of the full system.
In storage technologies which need DC supply, the PCS rectifies AC line power to
DC during charge, and inverts the DC power back to AC during discharge. The PCS
controls the rate of discharge and the switching time of the system. The power
switches are typically either GTO (gate turn off) or the newer, more flexible IGBT
(insulated gate bipolar transistor) semiconductors. IGBT semiconductors have fewer
requirements for driver circuitry, making inverters more compact and modular. IGBTs
are currently used to overcome problems of poor power factor and high current
harmonics and are employed in a number of wind turbine designs.
Additional PCS components may include transformers as needed for voltage matching
and isolation, and a controller for operating the system and interfacing with the
supervisory system.
The requirements placed on the PCS will depend on the applications that the system is
going to be used for. In application requiring a fast response, fast-acting power
conversion and control systems are as crucial as the dynamic behaviour of the storage
subsystem.
PCS for large-scale systems are not off-the-shelf components. The concept of modular
PCS is now being promulgated as a way to drive PCS costs down. Modular PCS is
comprised of many small converters that are networked in parallel to achieve the
same power rating of a single large converter, but benefit through the economies of
mass production. The individual units, if designed to operate with a sufficient degree
of autonomy, can be resealed dynamically. This offers the advantage of redundancy
81
Electricity Storage and Wind Energy Intermittency
Final Report
high efficiency at low power rates, because only the minimum required number of
power converters need to be energized.
The balance-of-plant includes structural and mechanical equipment such as the
protective enclosure, heating/ventilation/air conditioning (HVAC), and
maintenance/auxiliary devices. Other BOP features include the foundation, structure
(if needed), electrical protection and safety equipment, metering equipment, data
monitoring equipment, and communications and control equipment. Other cost such
as siting, permits, project management and training may also be considered here.
The system control is a key part of an electrical energy storage system, especially
when it is configured to match several different application concurrently. It must be
designed in a way that permits the realization of all these functions. An obvious
important consideration here is to ensure that the target applications do not impose
mutually exclusive demands on the storage system and when conflicting demands are
placed on the system that the highest priority need is met as Collinson (2000) points
out.
3.10 Costs
Table 3.7 contains a breakdown of all the possible expenditures of an electrical energy
storage system, including operation and maintenance.
Utility-scale storage systems are not at present off-the-shelf products (with the
exception of some power quality systems), and are custom-sized. The incidental costs
particular to tailored systems has added considerable cost to each of the systems now
in operation. Collinson (1999) stresses that modularity and standardisation of system
assemblies will reduce costs, improve reliability, and diminish the demands on system
designers and engineers.
The initial costs of an electrical storage system can be calculated by the sum of a notlinear cost function of the power rating, a non-linear cost function of the energy
capacity, and a fixed cost:
Initial cost: F(energy) + F(power) + Fixed cost
Schoenung (2001) groups the costs under three main headings:
• storage media cost (as a function of cost per kWh)
• power conditioning unit (as a function of cost/kW)
• balance-of-plant (as a function of cost/kWh)
Collinson (2000) follows a very similar line, although giving the BOP cost as a
function of the power rating (cost/kW). Indeed Schoenung acknowledges, that in
some cases it may be proportional to the installed power or even a fixed cost.
This approach provides a very rough basis for approximation. Although for some
technologies and applications the storage subsystem cost is clearly dominated by
energy-related costs, it is more a combination of energy-related and power-related
costs. This is especially true for PHES, CAES, FBES and HES, in which the charge
and discharge devices are independent from the storage media. In a HES system, for
82
Electricity Storage and Wind Energy Intermittency
Final Report
instance, the electrolyser and fuel cell costs are function of the power ratings. Even
for technologies in which there is not physical decoupling of power and energy
capabilities, such as BES, FES and SMES, the cost can be dominated by powerrelated costs in systems with a high power/energy ratio.
Storage subsystem
Interfaces to AC Load and
Source
Power Conversion System
Auxiliary
Accessories
Systems
and
Monitors & Controls
Facilities
Labour Costs
Operation and maintenance
New lines to serve installation
Transformer between utility voltage and battery system AC
voltage
Protection devices (e.g. switches, breakers, fuses)
AC switchgear/disconnect
Rectifier/inverter
DC switchgear/disconnect
Protection devices (e.g., switches, breakers, fuses)
Mechanical: racking/physical support, watering/heating/air
and fluid pumping systems
Safety equipment (e.g. ventilation, fire equipment,
detectors, respirators, spill troughs)
Cryogenic refrigeration, vacuum system
Monitors/diagnostics: storage media, power conversion,
subsystems (bearings, cryogenics, vacuum)
Controls: storage media, protection devices, power
conversion, subsystems
Foundation and structure
Materials
Lighting/plumbing
Access road and landscaping
Grounding/cabling
Heating, ventilation, air conditioning
Construction
Installation and start-up testing
Operations
Project management
Service contract: inspection, service costs, component
replacement
Training for operation and maintenance workers
Monitoring/data acquisition2
Financing, taxes and permits
Transportation
Table 3.7. Breakdown of the costs of an electrical energy storage system
Also, for some technologies, the cost is not linear over the range of sizes, due to the
economies of scale. This is not only applicable to the storage subsystem, but to the
PCS and BOP as well. There are many common elements that must appear in all
systems, and thus the cost does not scale linearly with size.
Operation & maintenance costs consist of fixed and variable costs. For instance, the
fixed costs of a battery system include cooling and general maintenance at the site,
while variable costs include recharging the batteries and periodically replacing the
batteries.
The contribution of each expenditure to the total cost will depend largely on the
application and technology. In general, the capital cost in power quality applications
is dominated by the power conversion system, followed by the BOP. On the contrary,
83
Final Report
Electricity Storage and Wind Energy Intermittency
in utility-scale energy storage the energy storage subsystem may prevail over the PCS
and BOP as illustrated in figure 3.18.
Utility-scale Energy Application
Power Quality Application
Other
Other
Energy
Storage Me
Energy
Storage Media
Balance of
Plant
Balance of
Plant
Power
Conversion
Equipment
Power
Conversion
Equipment
Figure 3.18 Examples of cost breakdown for different storage applications
The costs of the PCS can be very variable. Different outage protection features
would, for example, add different costs. The costs are not linear with the size of the
systems. While for a 300-500 kVA system the price of the PCS may be in the range
$300-235/kW, for a smaller system 30-50 kVA the cost would be as high as $500400/kW.
PCS has significant contribution to the overall cost, but they are becoming cheaper
and more modular, and the component ratings are increasing. Progressive
standardization is very important to achieve cost reductions, interchangeability from
different suppliers, and ability to replace them easily.
Standardisation is also crucial to bring down the costs of BOP. In mature
technologies, e.g. lead acid batteries, the largest potential for cost reduction is the
BOP. Again the cost of BOP is highly variable, and depends on a myriad of factors,
not only related the storage technology and the applications, but also the site for the
installation.
By way of example, the BOP cost of a large-scale BES system may be separated into
three components:
• Facilities to house the equipment: ~45%
• System design and integration: ~22%
• Transportation, finance charges and taxes: ~33%
Table 3.8 shows the contribution of the PCS and BOP in some large-scale BES
systems [see Akhil (1997)].
Plant
CHINO 10MWh/40MW
HELCO 10MW/15MWh
PREPA 20MW/14MWh
VERNON 3MW/4.5MWh
Storage
44%
34.5%
22%
32%
PCS
14% ($258/kW)
18.5% ($212/kW)
27% ($294/kW)
19% ($275/kW)
BOP
42%
47%
51%
49%
Table 3.8. Cost breakdown of large-scale BES systems
84
Final Report
Electricity Storage and Wind Energy Intermittency
3.11 Comparison of electricity storage technologies
The various storage technologies have a number of features, such as efficiency, size,
dynamic behaviour, etc. which are all relevant in the selection of the most suitable
system for a particular application or set of applications. Available power and energy
ratings are essential criteria for any application.
The energy / power ratings in Table 3.9, also illustrated in Figure 3.19, correspond to
the current realisations. Some technologies, such as the polysulphide-Br flow battery
are targeting power ratings far larger. Obviously, modular technologies such as HES
can theoretically reach unlimited size.
1.E+04
PHES
CAES
1.E+02
Energy (MWh)
HES
Discharge
time
Flow BES
1s
1.E+00
MetalAir
Hi-Energy
SCES
1m
Batteries
1h
10 h
1.E-02
1.E-04
Hi-P FES
SMES
Hi-Power SCES
1.E-06
0.01
0.1
1
10
100
1000
Power rating (MW)
Figure 3.19 Power and energy ratings of storage technologies
Life spans are often expressed in literature in terms of years. However, the
characteristics of the duty cycle affect the replacement interval of some components
and eventually the service life of the whole storage device. Similarly, the nature of the
duty cycle profile, its distribution and frequency also affect to some extent the
performance of virtually all storage systems. While frequent cycles increase the
efficiency of some storage media, they decrease efficiency of others.
Although not included in Table 3.9, the size of the systems (m2/kWh) may be a very
important parameter if space constraints exist. The portability may also be an
important factor in some applications, but for large systems requiring significant
energy content portable systems are not be feasible, as the size of the storage media
often becomes impractical or non-economic for transport. Portability varies greatly
between types of systems. SMES, battery and flywheel systems are now offered
commercially as pre-packed systems that fit into trailer containers with all of their
monitors, controls, and power conversion equipment for easy transportation and
installation.
The efficiencies in Table 3.9 correspond to the storage subsystem unless otherwise
stated. Power conditioning and parasitic losses need also to be considered.
85
Electricity Storage and Wind Energy Intermittency
Final Report
Regarding costs, the estimates presented in the previous review are simple ratios of
the total cost to the power or energy rating. However, with the specific costs given in
Table 3.10, the initial cost of a system can be estimated as an addition of powerrelated costs, energy-related costs, and BOP costs. Estimates can vary significantly
between different studies, and in most technologies costs are changing as they evolve.
86
Final Report
Electricity Storage and Wind Energy Intermittency
Pumped hydro
100 – 4000MW
Discharge
duration
4 – 12 h
CAES (in reservoirs)
100 – 300MW
6 – 20 h
sec – min
0.64
-
30 y
commercial
CAES (in vessels)
50 – 100MW
1–4h
sec – min
0.57
-
30 y
concept
Flywheels (low speed)
< 1650 kW
3 – 120 s
< 1 cycle
0.9
∼1%
20 y
commercial products
Flywheels (high speed)
< 750 kW
<1h
< 1 cycle
0.93
∼3%
20 y
prototypes in testing
Super-capacitors
< 100 kW
<1m
< 1/4 cycle
0.95
-
10,000 cycles
some commercial products
10 kW – 10MW
1s–1m
< 1/4 cycle
0.95
∼4%
30 y
commercial
10 – 10MW
1 – 30 m
< 1/4 cycle
0.95
∼1%
30 y
design concept
Lead-acid battery
< 50MW
1m–8h
< 1/4 cycle
0.85
small
5 – 10 y
commercial
NaS battery
< 10MW
<8h
n/a
0.75 – 0.86
5 kW/kWh
5y
in development
small
2,000 cycles
in test / commercial units
Power rating
SMES (Micro)
SMES
Response
time
sec – min
Efficiency
Parasitic losses
Lifetime
Maturity
0.7 – 0.85
evaporation
30 y
commercial
*
ZnBr flow battery
< 1MW
<4h
< 1/4 cycle
0.75
V redox flow battery
< 3MW
< 10 h
n/a
70 – 85*
n/a
10 y
in test
Polysulphide Br flow battery
< 15MW
< 20 h
n/a
60 – 75*
n/a
2,000 cycles
in test
< 250 kW**
as needed
< 1/4 cycle 0.34 – 0.40*
n/a
10 – 20 y
in test
< 2MW**
as needed
seconds
0.29 – 0.33*
Hydrogen (Engine)
*AC-AC efficiency
** Discharge device. An independent charging device (electrolyser) is required
n/a
10 – 20 y
available for demonstration
Hydrogen (Fuel Cell)
Table 3.9. Characteristics of storage technologies
87
Final Report
Electricity Storage and Wind Energy Intermittency
Capital cost
Powerrelated cost
($/kW)
600
Energyrelated cost
($/kWh)
0 – 20
425 – 480
CAES (in vessels)
O&M cost
BOP
($/kWh)
Fixed
Variable
Cost
($/kW-y) (c$/kWh) certainty
Environmental
issues
Safety issues
included
3.8
0.38
reservoir
exclusion area
3 – 10
50
1.42
0.01
gas emissions
none
517
50
40
3.77
0.27
gas emissions
pressure vessels
Flywheels (low speed)
300
200 – 300
∼80
-
containment
Flywheels (high speed)
350
500 – 25,000
∼1000
7.5
0.4
-
containment
Super-capacitors
300
82,000
10,000
5.55
0.5
-
-
SMES (Micro)
300
72,000
∼ 10,000
26
2
-
magnetic field
SMES
300
2,000
∼ 1,500
8
0.5
-
magnetic field
200 – 300
175 – 250
∼ 50
1.55
1.0
lead disposal
lead disposal, H2
259
245
∼ 40
n/a
n/a
chemical handling
thermal reaction
1,500
200
included
n/a
n/a
chemical handling chemical handling
n/a
175 –190
n/a
n/a
n/a
chemical handling chemical handling
1,200
175 –190
n/a
n/a
n/a
chemical handling chemical handling
Hydrogen (Fuel Cell)
1100 – 2600
2 – 15
n/a
10.0
1.0
-
-
Hydrogen (Engine)
950 – 1850
2 – 15
n/a
0.7
0.77
emissions
-
Pumped hydro
CAES (in reservoirs)
Lead-acid battery
NaS battery
ZnBr flow battery
V redox flow battery
Polysulphide Br flow battery
Price list available
Price quotes available
Cost determined each project
Costs estimated
Table 3.10. Cost comparison of storage technologies
88
Final Report
Electricity Storage and Wind Energy Intermittency
Chapter 4 STORAGE TO ACCOMMODATE WIND ENERGY IN IRELAND
4.1
Selection of technologies for different storage applications
Due to the fact that many different energy storage technologies can be used in a
variety of applications, identifying the technology which best matches the
requirements of an energy storage application can be a difficult task. Each storage
technology has particular characteristics that make it suitable in some situations, but
less desirable in others. It is important to identify the key features of different storage
media and to match the storage technologies with the most appropriate end-use
applications, including multiple-application systems, in order to achieve an optimum
cost-benefit solution. The selection process can be aided by a technoeconomic
cost/benefit model Collinson (2000).
Hydrogen engine
Hydrogen fuel cell
Flow batteries
Advanced batteries
Lead-acid batteries
Superconducting magnets
Supercapacitors
Flywheel
Compressed air
Pumped hydro
Storage Technology
Table 4.1 shows the technical suitability of the various technologies to the different
applications. It is based on a survey carried out by Schoenung (2001) with the
incorporation of flow batteries, which are potentially suitable for all storage
applications.
Storage Application
Transit and end-use ride-through
T&D stabilisation and regulation
Peak generation
Fast response spinning reserve
Conventional spinning reserve
Uninterruptible power supply
Renewable integration
Load levelling
Load following
Emergency back-up
Renewables back-up
Table 4.1 Technical suitability of storage technologies to different applications
The analysis carried out by Schoenung (2001) [and Schoenung (2002)] provides a
useful first indication of the technologies that are more suitable than others for
renewable integration. The study did not include flow batteries, which are widely
recognised as a technology with a significant potential for renewable integration and
89
Electricity Storage and Wind Energy Intermittency
Final Report
other applications. The costs for these storage technologies were those showed in
Table 3.10. Two cost estimates have been employed for the hydrogen storage option,
a low estimate based on a specific cost of $500/kW for the fuel cell and $300/kW for
the electrolyser and a high estimate, in which $1,500/kW is attributed to the fuel cell
and $600/kW for the electrolyser. Similarly, two different costs can be considered for
the lead-acid batteries according to the range expressed in Table 3.10.
The results of this cost comparison analysis are summarised in figures 4.1 – 4.4.
Figure 4.1 compares the capital costs of different technologies used for short duration
(< 2 hours) storage. For applications with discharging times ranging from minutes up
to 2 hours, BES and HES (low cost projection) seem to be the most competitive, with
CAES also likely to play a role in this area.
Schoenung (2001) also compared hydrogen-fuelled combustion engines and fuel cells
and these results are shown in figure 4.2. The engine appears a very competitive
solution for applications up to a few hours due to the lower capital costs. At longer
discharge times the lower efficiency begins to dominate. In addition, engines exhibit a
slower dynamic response.
Figure 4.11 Costs of applications <2 h
discharge
Figure 4.11 Comparison of batteries
and hydrogen technologies <2 h
discharge
As the discharge times increase, extending the applications to load management, the
power related costs (per kW) become less important and the energy related cost (per
kWh) begin to dominate, as evident from table 3.10.
Figure 4.3 compares the capital costs of storage technologies that are technically
suited to long duration storage. Two traditional technologies, PHES and CAES are
least costly but these do have siting limitations however. The next most attractive on a
capital cost basis is the hydrogen fuel cell (low cost scenario), and then CAES storing
air in tanks. Lead-acid batteries seem to lose attractiveness for long discharge
applications, but this is very sensitive to cost projections.
Fig 4.4 show that HES costs exhibit a lower cost increase rate than lead-acid batteries
as the energy storage capacity grows, but the latter would be still desirable if the
hydrogen storage components remain expensive.
90
Electricity Storage and Wind Energy Intermittency
Final Report
Figure 4.12 Costs of applications for long duration storage
Figure 4.13 Comparison of batteries and hydrogen fuel cells. Long duration
For even longer storage time (several days), Schoenung (2001) points to PHES, HES
and CAES as the most economic options. It should again be noted however, that flow
batteries are not included in the study.
4.2
Experience in the use of storage for the integration of RE
There is little or no experience on the use of electrical energy storage for the
integration of large-scale intermittent renewable generation on electricity systems.
In Germany, there have been some multifunctional energy storage systems used to
improve the utilization of renewable energy supplies (Wagner, 1999). These systems
include three different functions: UPS, improvement of power quality, and peak-load
shaving. For such a multifunctional application, large lead–acid batteries with high
power and good charge acceptance, as well as good cycle life was deemed
appropriate. One system was installed in combination with a 2MW/1.2MWh wind
91
Electricity Storage and Wind Energy Intermittency
Final Report
farm in Bocholt. The batteries were modified according to the demand of a
multifunctional application, and an appropriate power converter was designed. The
use of a battery made it possible to substitute the required peak-load power with the
stored power from renewable energy sources achieving high efficiency.
As previously mentioned, Tennessee Valley Authority (TVA) had plans to install a
Regenesys storage system parallel to the development of a 20MW wind farm. The
main reason for the initial decision of building the plant was that in the TVA’s area
the wind resource is most available at night during much of the year. In contrast,
TVA’s greatest needs for intermittent or peaking energy are during the morning in the
winter, and during the afternoon and early evening in the summer. TVA was hoping
that the system would meet peaking needs, improve power quality and reliability, and
provide rapid response to changing power demand. TVA decided against the project
in 2002 and in December 2003, RWE Innogy ceased the development of the
Regenesys technology.
ZBB is targeting the renewable integration market for its ZnBr flow batteries, which
can be used to smooth out the wind farm fluctuations and hence help to control the
frequency fluctuations. The company is in negotiations with Apollo Energy
Corporation to provide 30 ZnBr batteries to back up a 20MW wind farm for several
minutes. The goal is to keep the wind farm operational for the 200+ hours each year
when erratic winds would otherwise force operators to shut down some turbines.
Pumped Hydro storage systems have been proposed to increase the wind penetration
in the Greek islands of Crete [Christakis (2001)] and Serifos [Theodoropoulos
(2003)]. As pointed out in Section 2.4, the Crete system is the largest isolated or
‘island’ electricity system with wind input. Given the weakness of the
interconnections in Ireland, Crete might resemble the future Irish situation more than
Denmark.
The major difficulty for the system operator in Crete is that a minimum quantity of
conventional generation must be kept operating, to provide frequency control and
reactive power. This restriction leads to the curtailment of the wind production at
times of high winds and low demand. The level of curtailment in 2001 was 6% of
wind production, and was expected to rise to approximately 20% in 2002, due to
additional wind farms coming onstream [see Garrad Hassan et al (2003)]. This has
clearly a major effect on the economics of wind generation in Crete. Studies show that
the PHES units can contribute a lot toward the minimisation of the existing thermal
units energy production cost [Christakis (1996)] and the maximisation of the wind
energy penetration [Christakis (1997)].
Zaininger (1997) examined the benefits and costs of installing an integrated MWscale wind farm with battery storage to defer the upgrade of a 25 kV circuit to 69 kV
for Orcas Power and Light Company in US. Although sufficient wind potential was
identified, the high winds did not generally occur coincidentally with peak loads on
the distribution line. A transportable 500 kW/2-hour battery was considered for use
during low wind periods to defer the upgrade of the distribution line until 2000.
Enslin (2004) explored energy storage options in the context of Dutch plans to
connect 6000 MW of offshore wind farms off the Dutch coast. The study considered
92
Electricity Storage and Wind Energy Intermittency
Final Report
PHES, CAES, Regenesys and lead acid batteries and concluded that storage would
not be an economically viable option, if only considered for one function and should
be considered for several parallel tasks, including power balancing, grid stability,
power trading and power quality mitigation solutions.
Gonzalez et al (2003) made a preliminary examination of the role of hydrogen in
facilitating high wind energy penetration in Ireland.
4.3
Assessment of different technologies
In assessing technologies it is important to bear in mind that electricity market
policies and regulations will influence the choice of storage technology best suited to
renewable energy integration. Environmental policies will also have an impact and are
likely to disfavour options posing environmental hazards. It is essential to bear in
mind that future regulations could affect the chosen technologies in a positive or
negative manner.
The rate at which the technologies develop is critical to their future attractiveness. The
achievement of certain milestones, especially in terms of costs, can change the best
option.
An ideal storage technology candidate for renewable energy applications would have
the following characteristics:
•
Low cost, long cycle life and little maintenance requirements
•
Mature technology
•
High efficiency
•
Large energy storage capability
•
Flexibility and adaptability to future trends in the electric sector
•
Modularity and easiness of upscale
•
Ability to deliver power rapidly or slowly, as desired, under full control
•
Ability to operate on a reversible charge/discharge cycle. Capability of deep
discharge without damage.
•
Environmental sustainability and safe operation
•
High energy density in some cases
•
Possible synergies
None of the technologies described in section 3 meets all these requirements.
The cycling requirements depend on the functions that the system is going to fulfil.
For instance, if used for load levelling, the system would typically go through a
discharge/charge cycle once or twice a day, mainly during weekday peaks. If storage
is used to adjust the output to the scheduled generation, the cycling would depend on
the mismatches between forecast and actual renewable power output.
93
Electricity Storage and Wind Energy Intermittency
Final Report
Pumped hydro
Even though PHES can provide high storage capacity at low cost, the application to
renewables is constrained by a number of factors. A crucial limitation is the siting
requirements (large water reservoirs), which inhibits further developments. PHES is
normally used at large scale level by major utilities, because of the time and capital
needed to construct them. The scale is not particularly suited to distributed generation.
Its main applications are energy management, frequency control and provision of
reserve. Its response is rapid enough in both directions for these applications but
limits its deployment for renewable integration, since it cannot follow wind energy
variations rapidly. In large systems where quick response is not necessarily required,
it is suitable for bulk renewable storage and seasonal storage. Due to these
considerations, Herr (2002) concludes that PHES is not likely to be employed in
widespread renewable integration, even with the offer of low cost storage capacity.
Compressed air
CAES is the only commercially available technology other than PHES able to provide
very large energy storage deliverability (above 100MWh) to use for commodity
storage or other large-scale setting. Operational experience is very limited however,
as only a few facilities have been installed worldwide to date. The response time is
another drawback. According to Gordon (1995), CAES appears to be the most
economic option for systems that require 3-12 storage hours per day.
Smaller-scale installations could have less stringent siting requirements, shorter
construction times and require moderate investments, although the specific costs
prove higher. These micro-CAES systems could be integrated at the distribution level.
In 2001, the US mining company Ovoca Resources announced that it had entered into
a joint venture arrangement with the purpose of making a preliminary feasibility study
of compressed air energy storage possibilities in Ireland. Results of this study
indicated considerable potential and this led to the decision of setting up a joint
venture, Optimum Energy Limited, owned by Ovoca and Mercury Holdings.
Optimum Energy has identified the storage of electrical energy for use at peak times
as a highly profitable and yet undeveloped aspect of the electricity market in Ireland.
In addition to the on-peak/off-peak differential, Optimum sees storage as a key
component in the strategic development of wind energy in Ireland. In this regard, they
claim the system to have fast reaction times and could reduce largely the need to hold
hydrocarbon-powered plants on spinning reserve. The intended storage vessels for the
compressed air are deep underground areas of high rock porosity (several natural gas
storage plants use similar underground structures). Optimum reports that following
detailed analysis of existing geologic and drilling data, a number of potential sites
suitable for CAES development have been selected, although the project's full
feasibility, particularly in relation to reservoir integrity and suitability, is yet to be
proved. This will involve geotechnical surveys, drill testing, computer modelling, and
so on. Optimum budgeted a sum of up to €1.3m for this phase.
94
Electricity Storage and Wind Energy Intermittency
Final Report
Flywheels
FES systems are in early stages of market entry and are primarily expected to serve
the customer-end power quality market. Whereas steel-rotor flywheel has very limited
scope for the entire array of applications, composite-rotor FES has potential for
broader applicability. However, it will require significant development to compete
with other, more mature technologies and non-technology options
Application of FES to RE integration is under consideration. Although they are
unsuitable when large amounts of stored energy are required and storage and
discharge times grow larger, further improvements in storage capacity would make
them more suitable for renewable integration. The use of a number of small units in
parallel can meet large energy storage needs, but this solution does not benefit of
economy of scale.
Since composite flywheels depend primarily on high speed to achieve the necessary
power and energy levels, they also depend on high-strength fibres to allow for
lightweight, high-speed rotation. Unless the cost of the fibre material (advanced
carbon fibres) falls significantly, Butler (2002) points out that FES systems will be
limited to applications with low duration discharge.
However, in weak grids FES can also help to comply with the grid requirements,
avoid grid reinforcements, and facilitate higher penetration rate (improvement of
power quality by dynamic absorption/injection of reactive power). In remote areas,
FES can improve the power quality, the overall efficiency and the durability of hybrid
systems.
In smoothing out short and medium-term variations, FES is superior to BES due to
the cycling limitations of batteries. As, conversely, FES cannot compensate for longperiod intermittency of the wind resource, it has been proposed that flywheels and
could used in combination with batteries, each addressing different challenges
associated with renewable integration. However, this increases the costs too much.
Indeed costs are an important issue for FES, although the long life and minimal
maintenance are compensating factors.
Most flywheels manufacturers are targeting the distributed power generation market,
with special focus on renewable sources.
Supercapacitors and supermagnets
Both SCES and SMES devices are being developed primarily for grid stabilisation,
uninterruptible substations and pulse power applications and their use for renewables
integration is not the immediate current focus. They are unsuitable for long-duration
storage due to the high capital costs per kWh. In SMES, the high energy consumption
by the cryogenic and refrigeration systems is another disadvantage for long-term
storage. One application where they could make their mark in time, however, is in
managing the power quality of wind farms. The stabilisation of transmission and
distribution lines increases the capacity of the grid to accommodate wind energy.
95
Electricity Storage and Wind Energy Intermittency
Final Report
Batteries
Traditionally, the use of BES in utility applications has been hindered by uncertainty
about lifetime, lack of understanding on how to apply battery systems, lack of
operating experience, maintenance needs, reliability issues, and desire to reduce initial
investment.
However, Butler (1996) reports that the experience gathered from early battery
installations such as the BEWAG and Chino plants, has generated a positive reaction
from the utility community. As a result, both LA and VRLA batteries are in an
advantageous position. At any rate these technologies have the most field experience
and can satisfy most of the defined utility energy storage applications, unless footprint
is important. LA batteries have a long record as renewable buffers in small off-grid
systems. Although a mature technology, still some improvement can be expected with
regard to energy capacity, lifetime, and recharge times. LA and VRLA batteries,
though affordable initially thanks to economies of manufacturing scale, are less so in
the long term because their limited cycle lives.
As for NiCd batteries, the plant being built in Alaska, which will provide spinning
reserve, shows the sort of scale of storage regarded as viable for some renewable
applications. However, the high costs and the environmental issues related to
cadmium disposal make this technology little desirable for renewable integration.
Advanced and flow batteries are considered potential candidates for all applications,
but have not yet been fully evaluated, and some of the technologies are just emerging
as pilot-scale systems. Lithium-based batteries, such as lithium-ion and lithiumpolymer are still only available in small sizes. Scaling up of these technologies for
utility applications is not likely to happen in the medium term.
NaS batteries are often considered a potential candidate in renewable application
[Butler (1996)], although there are contradictory positions in this regard [Butler
(2002)]. In theory, the power and energy ratings achieved in current realisations make
NaS suitable for medium-scale renewable integration. However, these batteries are
still at an early stage of technological maturity. Costs need to be reduced and service
life extended to compete with other technologies. Another hurdle is the high
temperature operation, which requires thermal management and involves parasitic
losses.
Metal-air batteries may become an option in the long term on account of their
potentially very low cost and high energy capacity, but they are at an early stage of
development and the efficiency is very poor at the moment to consider them as an
alternative.
Flow batteries
According to Collinson (2000), FBES systems are the best option for long-term
storage. Their main advantage is the decoupling of power and energy ratings. The cost
of additional storage capacity is limited to the active materials and storage tanks. FBs
can provide both power-intensive and long discharge, and can cycle very rapidly and
deeply.
96
Electricity Storage and Wind Energy Intermittency
Final Report
Thanks to the rapid response and the low operating costs when idle, flow batteries are
well suited to supply ancillary services such as voltage and frequency regulation and
spinning reserve according to Taylor (2002). Therefore the versatility of FBs makes
them suitable to provide a number of secondary services at the same time as matching
or levelling load and generation. This offers critical advantages, but in order to
improve FBES competitiveness with respect to the cheaper, proven systems such as
LA and VRLA batteries, it is essential that these multiple services are commercially
developed and proven in the marketplace.
The life cycle of FBs is longer than classical batteries, offering life cycle cost
advantages, but this claim is based on relatively little experience. As a new and
unfamiliar technology, any FB successes that build the market’s confidence will
benefit all FB producers. Analysts project continued FB initial market penetration
over 1–3 yrs, and growth to commercial-scale production over 3–5 yrs according to
Lotspeich (2002).
FB technologies are evolving as they enter the marketplace. Orders are few, current
capital costs are high, and comparisons between firms and technologies are neither
simple nor direct. Technical characteristics are similar, with no system offering
clearly superior performance, according to Lotspeich (2002), who provides a more
detailed comparison. This comparison is summarised in table 4.2 and was made
before the ending of the Regensys programme in December 2003.
Commercialisation characteristics
The three major FBs have relative attributes and commercialisation pathways that
help define particular markets where each might compete more or less effectively.
Regenesys focuses on multi-MW systems roughly tenfold larger than typical VRB
and ZnBr FBs. Each Regenesys® installation is to be built as an integrated, turnkey
system, with scalable electrolyte storage. This might reflect a focus on facility scale
economies of power and energy capacity. Production capacity was not determined,
but Regenesys is reportedly well positioned for manufacturing according to Lotspeich
(2002). VRB and ZnBr FB developers are concentrating on modular systems,
typically (but not exclusively) below 1MW of power capacity. This might reflect a
focus on production scale economies for FB and auxiliary system components. In
2000, VRB firms produced 10MWh compared to ZnBr production of 4.5MWh. Some
VRB and ZnBr firms are also targeting larger-scale markets, where they would
compete with Regenesys®. Regenesys® has the potential to compete best in multiMW applications, e.g., power trading, large generation and transmission and
distribution (T&D) grid support; VRB markets span from generation and T&D grid
support to facility-scale applications, while ZnBr FBs might compete best in facilityscale and distribution- or substation-level support.
Economic comparison
Power and energy capacity costs are useful but vary considerably between different
applications and systems. Life-cycle cost of ownership is arguably the most useful
metric. Cost data varies and is hard to get from competing developers, but some
information is available. A study carried out by the US Electric Power Research
Institute (EPRI) in 2000 evaluated total costs for VRBs, Zn-Br FBs, and Regenesys®
97
Final Report
Electricity Storage and Wind Energy Intermittency
Vanadium redox
Na polysulphide / Na bromide
Zinc / Bromine
Net efficiency (AC – AC)
~70–85% depending on operation
55–75% depending on operation
~75%
Lifetime
Cycles: ≥10,000
Target: 10 y; 15 with O&M
Cycles: n/a
Target: 15–20 y with O&M
Cycles: ≥ 1,500 cycles
Target: 10–20 y with O&M
Power cost
$1,500–5,500/kW
Projected: $1,000/kW
~$1,500/kW
Projected: ~$750/kW
$1,500–2,000/kW?
(ZBB: ≥$800/kW achievable)
Energy cost
Electrolyte: $30–50/kWh
System: $300–1,000/kWh
Electrolyte: $10–20/kWh
System: $160–185/kWh
Electrolyte: ≥ $10–20/kWh
System: target ~$400/kWh
O&M cost
~$50,000/y for 2.5MW, 10MWh
n/a
$30,000–150,000/y for 2.5MW,
10MWh
System cost
~$11 million for 2.5MW, 10MWh
~$300,000 for 100 kW, 100 kWh
$20–25 million for 10–15MW, 100–
module; $5.8–8 million for
150+MWh
2.5MW, 10MWh
Representative systems
250 kW, 520 kWh;
1.5MW, 1.5MWh
12–15MW. 120MWh
Projected capacity
50 kW, 500 kWh to 5MW, 20MWh;
5–50MW, 100–250+MWh;
50–100MW upper range; 500MW
500MW feasible
feasible
50 kW, 500 kWh module;
200 kW, 400 kWh trailer
300–600 kW, 300–1,000 kWh
modular arrays; 4–5MW, 4–
10MWh upper range
(“no practical limit”)
Table 4.2 Comparison of the three FB technologies (Lotspeich, 2002)
98
Electricity Storage and Wind Energy Intermittency
Final Report
for a large system (10+MW, 100+MWh). TVA (2001) used the study to select its
Mississippi FB system, and summarized the results. Sandia National Laboratories
(SNL) and Black&Veatch (B&V) (2001) surveyed Vanteck, SEI, ZBB, and Powercell
to evaluate technical and cost factors for a 2.5MW, 10MWh battery demonstration
project in Nevada .
POWER CAPACITY COSTS: FB cost/power ratio varies by application and
discharge rate, complicating direct comparisons. The three FBs are in a roughly
equivalent range of a few to several thousand dollars per kW for initial systems,
decreasing towards $1,500–2,000/kW (ZBB reports that ~$800/kW and lower is
achievable). TVA (2001) reports that EPRI concluded that Regenesys® had the
lowest capital costs of the three main designs. VRBs and Zn-Br FBs developers focus
on smaller, modular systems, which might reduce capacity costs faster as production
volumes build over the mid- to long-term. SNL/B&V indicated that ZnBr systems
offer lower capacity costs than VRBs.
ENERGY CAPACITY COSTS: As the energy capacity increases, the cost/energy
ratio is more influenced by the electrolyte costs, which are not consumed during
cycling. Lotspeich, (2002) points to analysts estimating electrolyte costs in the range
of $10–20/kWh for Regenesys® and ZnBr FBs, and $30–40/kWh for VRBs. EPRI
indicated that ZnBr electrolyte costs per kW were twice that of Regenesys®.
Regenesys® appears to have provided the lowest system electrical energy storage
cost, reportedly in the range of $160–185/kWh per kWh in large installations. ZBB
reports ZnBr FBs storage costs approach $400/kWh with scaled-up production.
Consistent values for VRBs were not determined. Apparently ZnBr systems currently
offer lower electrolyte and total capacity costs than VRBs. However, VRB systems
offer greater scalability of electrolyte storage, probably enabling storage cost
reductions in larger installations that stacks of electrically linked but not hydraulically
connected ZnBr modules cannot attain.
PCS AND CONTROLS COSTS: PCS costs are typically $200–400/kW for smaller
(~100 kW) batteries, and are roughly equivalent components for all FBs on a capacity
and cost basis according to Lotspeich (2002). Although controls and PCS for energy
storage technologies are widely available, FB firms are developing proprietary
components, software, and integrated systems to match FBs’ diverse capabilities, and
anticipate cost and performance improvements. One VRB developer projects eventual
PCS cost reductions of 30–50%.
O&M COSTS: Long-term operating and maintenance costs are projected as FBs are
new and existing installations vary in design, capacity, and operational profile. All FB
systems are designed for automated operations, but initial installations are provided
with more maintenance and operational support. Regenesys® might require more
regular maintenance (e.g., for removal of process by-products) than VRBs or Zn-Br
FBs. Regenesys® installations target a 15 y service life, but the O&M cost to attain
that was not determined. Sandia National Laboratories and Black & Veatch (2001)
report average projected O&M costs for VRBs of ~$50,000, lower than the ~$90,000
of projected ZnBr costs for an equivalent capacity. VRBs are targeting a shorter
service life (7–15 y) than ZnBr systems (10–20 y). However, SEI claims to have
demonstrated cells that exceed 10,000 cycles and suggests a 10 y VRB stack service
99
Electricity Storage and Wind Energy Intermittency
life of their VRB, while Powercell offered buyers a 5 y service guarantee and
estimated only a 1,500 cycle life for its ZnBr modules.
TOTAL COST: The TVA study concluded that Regenesys® “would most likely
provide the lowest cost of operation of a life-cycle-cost basis for multi-hour utility
energy storage, while providing other energy storage services that have economic
value to electric utilities.” Zn-Br was the second choice, but the Zn-Br developer was
focused only on units for “short duration, small scale discharges (25 kW for 4
hours).” The study also concluded that VRBs and Zn-Br FBs had lower power and
energy ratings and higher capital costs than Regenesys®, and that Zn-Br “electrolytes
that are twice the cost per kilowatt of those used for Regenesys.” Table 4.2
summarises the features of the three technologies.
Activity in the RE sector
VRB in the only of the three technologies boasting a demonstration plant associated
with a wind energy installation, although of modest scale. A 170 kW, 1.2MWh
system is being used to stabilise the output of a 275 kW wind turbine operated by the
More noteworthy, Hokkaido Electric Power Company. Hydro Tasmania has awarded
recently a contract to Pinnacle VRB for the supply and installation of a VRB as part
of the King Island (Australia) wind farm expansion.
The ZnBr FB manufacturer ZBB is targeting the renewable integration market with a
module especially designed to operate in connection with renewable sources. Some
ZnBr units have been installed in off-grid systems in connection with solar
generation, but there is no experience with wind energy as yet.
As mentioned before, Tennessee Valley Association (2002) evaluated and selected a
12MW Regenesys® energy storage facility to operate in conjunction with a 20MW
wind farm, principally for load levelling. However, TVA (2002) eventually
abandoned the plan. Other energy storage technologies were rejected based on their
life-cycle costs, environmental impacts, and energy storage capacity. TVA was at that
time already constructing a Regenesys® facility in Mississippi to demonstrate its use
in meeting peaking needs, in improving power quality and reliability, and in
providing rapid response to changing power demand.
It is worth mentioning the reasons TVA argued to select Regenesys® in an
assessment carried out in collaboration with the EPRI. TVA was seeking a technology
in the process of being commercialised which could provide the lowest cost of
operation on a life-cycle basis for multi-hour utility energy storage while still
providing other services. TVA (2001) felt that PHES and CAES were found to
require large sites that are seldom available near the point of need and have negative
environmental and financial impacts. Included in this study were VRB, and ZnBr
FBs. Regenesys® was identified as meeting all of these requirements. ZnBr FB
ranked as the second choice, but with the drawbacks of being only developed for
shorter duration (< 4 h), small-scale discharges, and a higher cost of the electrolytes.
At the time of the report, the developer surveyed (probably Powercell) was focusing
only on this small-scale unit with no plans for larger scale plants [TVA (2001)].
100
Electricity Storage and Wind Energy Intermittency
Hydrogen
Due to low round-trip efficiency, hydrogen may not be best suited for renewable
integration unless very long storage durations are required.
As pointed out before, hydrogen role in the integration of RE is likely to be as a
source of fuel for other applications, mainly transport. At least in the medium term,
transport is believed to represent a higher market value for hydrogen than stationary
electricity generation. But even for this purpose, technological advances and cost
reductions are still necessary.
The production of hydrogen from wind generated electricity for transport applications
alone does not provide all the possible benefits for the network. The same level of
spinning reserve must to be kept, and the capability of stabilising the network is much
more limited. Hydrogen production, however, does contribute to smoothing out
considerably the wind power output, thus enhancing the electric system management
and avoiding curtailments.
Operational penalties when the demand exceeds generation would still stand and the
precise details of these in the new market arrangements are not yet fully clear, but it is
unlikely that in the medium-term the penalties are high enough to make regeneration
cost-effective.
The feasibility of the production of hydrogen for transport applications will be
subordinated to the success in building a hydrogen supply infrastructure and
commercialising hydrogen-fuelled vehicles. In that future scenario, hydrogen will
have to compete against the production of hydrogen from natural gas, which is for
now the most affordable process. Hydrogen is therefore an alternative envisaged for
the long term. Nevertheless, the route towards the hydrogen economy will require
intermediate steps, in which moderate amounts of hydrogen will be needed to supply
fuel to fleets of cars.
4.4
Conclusions
PHES viability is subject to siting requirements and environmental policies. The
construction of pumped-hydro facilities with enough capacity to address the growth
of wind energy in Ireland could face stiff opposition, even if suitable locations are
found.
CAES is a very promising solution provided that suitable underground caves exist in
Ireland. Although there are positive reports, an independent study of the potential in
Ireland is necessary. CAES will be probably the most cost-effective large-scale
energy storage technology, at least in the short and medium term.
FES may play a role in smoothing out short-term variations of wind farms on weak
distribution networks. They can also provide some services such as reactive power
compensation and frequency control. FES may increase the capacity to accommodate
distributed wind energy, but significant technological advances and cost reductions
are still necessary. At any rate, due to the limited energy storage capacity, FES can
101
Electricity Storage and Wind Energy Intermittency
barely address the loss of value of wind energy at high penetration, and therefore its
potential to allow for a large penetration of wind energy in Ireland is very
constrained. Similarly, the role of SMES and SCES would be limited to the
stabilisation of transmission and distribution lines, thus also enabling the system to
assimilate more wind energy.
Classical and advanced batteries would not be suitable when large amounts of stored
energy are required (costs largely proportional to energy capacity). BES advantages
over FBES seem to be limited to the greater operational experience and lower initial
costs. But even these advantages are expected to fade in the near future as FBES gain
market presence. BES might find applications in the short to medium term for
transmission and distribution capacity deferral, as these lines become congested at
times of strong winds.
Among the FBES technologies, the typical scales remain complementary. With the
demise of Regenesys®, which had the lowest cost of ownership, and was focused on
applications of at least 5MW (5 to 500MW systems were envisaged), the remaining
contenders are VRBs and ZnBr FBs. ZnBr FBs apparently have lower power and ES
capacity capital costs than VRBs, but VRBs offer higher efficiency and more scalable
storage. Both can potentially be suited at an individual wind farm scale to addressing
intermittency concerns.
As for hydrogen, the production of hydrogen using surplus wind energy will be of
serious interest in the short to medium term Specifically benefits will accrue to wind
energy producers operating in a liberalised market where they must rely on their own
ability to price electricity for dispatch and confirmed supply. Ultimately the wind
energy configuration of hydrogen to production of energy may become substantially
decoupled. This will allow for remote hydrogen storage and energy regeneration. The
key to hydrogen current viability is the nature of its solution of the intermittency
problem, and the effect of this on the realised price for wind hydrogen energy.
The options that will be analysed in the next section for the integration of growing
amounts of wind energy in Ireland will be: CAES, Regenesys and Hydrogen.
102
Electricity Storage and Wind Energy Intermittency
Chapter 5 ECONOMIC VIABILITY OF STORAGE OPTIONS
5.1
Introduction and summary
This chapter examines the economic viability of energy storage in a systematic
fashion. A number of studies have examined the effects of scale and technology
efficiency on the viability of energy storage systems. [Kroon (2002), Liu( 2003), EA
Technology (1998), EA Technology (1999) Sandia (1994) and Sandia (2002)]
Given these inputs, the preceding discussion on the appropriate technologies for
addressing intermittency, it has been possible to analyse the trends in terms of
research developments and economics, and to model the resulting outcomes in an
outline fashion. The consensus view among researchers points to significant expected
benefits from new technology developments (particularly scale, capital cost, charge
time, discharge time and efficiency), which will mirror those efficiencies currently
being achieved in wind energy production.
System prices for a range of energy storage solutions are likely to fall in the longer
term to below € 1,000 / kW as capital cost, for charge, storage and discharge
combined [Price (2000), Nakhamkin et al (2001), Baker (2001), Carnegie (2001) and
Harrison (2003)]. At the same time, progress has been made in decoupling storage
capacity, charge capacity and discharge capacity. Price data indicate that ideally in a
liberalised market, prices will exceed € 30 /MWh for periods most days, and € 50
/MWh during high demand periods. Effective storage technologies will allow
renewable generated electricity to sell into the market at these attractive prices. [Platts
(various)]
Energy storage, its interaction with liberalised electricity markets, renewable energy
systems, and the evolving requirements of electricity grids (allowing for
simultaneously increased local distribution and regional interconnection), is a
complex and fast growing research and industrial area. There are no absolutes in this
field, particularly with reference to setting the limits of emerging technology, and
market dynamics. [Garrad Hassan et al (2003) and ESB NG (2003a)]
Industry participants should be aware of the pace of research and development in the
energy storage field. Researchers should also take care to develop a complete cost
and economic profile, and to include direct, indirect and external costs in their
analysis. The analysis is, by nature, inter-disciplinary and complex, but is particularly
important with reference to the integration of renewable energy into power grids
designed for conventional generating equipment.[Liu (2003) and Garrad Hassan et al
(2003)] At the same time electricity markets are currently being liberalised (as is the
case in Ireland), thus increasing the complexity of the analysis. [EA
Technology(1999) and Kroon (2002)]
As discussed in chapter 4, the most likely systems to be appropriate in the Irish
context are:
1. Pumped hydro
2. Compressed air storage
3. Flow batteries
103
Electricity Storage and Wind Energy Intermittency
4. Hydrogen systems
In monetary terms, an energy storage system, which delivers energy at a cost of
5.1c/kWh (5.1 euro cents per kilowatt hour) or less, are closest to economic viability.
Energy storage systems based on pumped hydro and compressed air, can meet this
requirement now. Flow battery systems are expected to meet this cost requirement in
the short term, and large scale wind hydrogen systems have potential to achieve this
cost basis in the medium to long term (5-10 years). Table 5.1 gives comparative
measures for this target energy cost in terms of comparable primary sources:
Energy
Source
Energy
Density
Kgs to Provide
1MWh
Target Cost
Per Kg to give
€ 51/MWh
Coal
5.7 kWh/kg
175
€0.3
Natural Gas
9.4 kWh/kg
107
€0.5
Oil
11.4 kWh/kg
87
€0.6
Hydrogen
39.4 kWh/kg
25
€2
Table 5.1: Comparative Target Costs per Kg for Primary Energy Sources
Typically Combined Cycle Gas Turbine (CCGT) competes with energy storage
(particularly pumped hydro) in terms of ensuring power quality to a transmission grid.
From a technology perspective, Compressed Air Energy Storage (CAES)
complements CCGT systems and is not a competitor. Similiarly CCGT systems do
not compete with energy storage in applications where the storage system targets the
integration of wind energy to a transmission grid [Schoenung (2002)].
This study has analysed in some detail the challenge to provide this integration
capability in a technologically sound and economically viable fashion. The
fundamental report from this analysis is that large scale energy storage could be
economically viable in the medium term and would be well suited to the proposed
changes in the Irish electricity market, where renewable energy is not likely to be
preferentially treated in the future, where the grid is isolated currently and where
there is large scale potential for the development of wind energy [ESB NG (2003a),
CER (2003c)].
An alternative solution to the intermittency of wind power, the provision of a
‘supergrid’ capability was not evaluated. A supergrid would be capable of delivering
power to temporarily calm areas of Europe from windier areas. The underlying
assumption that there is a low incidence of continental calms has not been tested.
Summary economic findings:
In summary the findings of a preliminary but inclusive economic evaluation bear out
the following points:
1.
Grid requirements for power quality, power consistency and dispatch
prioritisation, for high baseload conventional plant, provide a rationale for
104
Electricity Storage and Wind Energy Intermittency
2.
3.
4.
5.
6.
7.
8.
energy storage. The economic case for these ancillary services is well
established and holds true in the absence of either a liberalised market or large
scale penetration of intermittent renewable energy [Price (2002)].
Electricity storage technology even decoupled from renewable sources is a
viable investment in an environment of a liberalised trading electricity market
[Kroon (2002)]
Future improvements in electricity storage technology are likely to further
improve energy storage’s exogenous attractiveness [Baker (2001)]
Electricity storage technology will allow integration of renewable sources to
replace conventional generation technology in an interconnected grid, thus
allowing a more aggressive penetration than a ‘fuel saving’ only analysis could
support [Garrad Hassan et al (2003) and Christakis (1996)]
Similarly a grid incorporating high renewable penetration and active storage
technology may not have to resort to wind ‘curtailment’ in any but the most
extreme circumstances. A corollary of this is that ‘curtailment’ usage such as
desalination or large industrial projects will likely not be appropriate. Ultimately
non-dispatched wind power will always be stored. [Liu (2003) and DOE (2003)]
Large scale renewable penetration, associated storage and ‘curtailment’ of
conventional generating capacity is a technical and financial possibility. The
economics of such an integrated system will likely depend on progress of
hydrogen wind systems, which are the only renewable systems with long term
discharge characteristics.[ Liu (2003), DOE (2003)and Lyons and Voigt (2003)]
TSO’s (transmission system operators), DSO’s (distribution system operators),
as well as regulators should understand the challenges and opportunities offered
by a large scale development of renewable energy linked to energy storage
systems.[ESB NG (2003a)]
The proliferation of storage systems (e.g Wind Hydrogen, CAES Hybrids,
Pumped Hydro, Advanced batteries), and of systemisation technologies are an
opportunity for a knowledge economy like Ireland’s to win a leadership role in a
new and potentially valuable emerging industry.[Dept. of the Taoiseach (2003)]
There is much interest within the EU 6th Framework Programme on energy storage.
Research opportunities provided by the both EU, National and private funding should
be encouraged. Ireland has produced comparatively little research to date on the
subject of energy storage. Opinions have often been formed based on limited
understanding of the emerging landscape, where research and development
expenditures are proliferating. The comparative low cost of production and capital
expense of Combined Cycle Gas Turbines (CCGT) generating equipment has further
entrenched a view that energy storage is not viable. However as pointed out above
this view lacks validity, as the core comparison to be made concerns the underlying
feasibility of technology and investment to reduce the intermittency of wind, and to
improve its dispatchability to the grid. There is need for further work on economic,
regulatory, scientific and engineering sectors in this field, particularly to understand
the emerging economics of the sector.
105
Electricity Storage and Wind Energy Intermittency
5.2
Costs: Present and Future
Economic Underpinings
In a fixed feed in tariff, or under a high AER priced contract, or similar systems in
Germany and elsewhere, there is little incentive to the renewable producer to store
energy. However under a system where renewable producers participate in the
market periods occur when low prices make it appealing to store energy rather than to
sell into the market at low or uneconomic rates. This will be the case under the
proposed incorporation of renewables into the market arrangements for electricity in
Ireland, and the UK approach of ROC (renewable obligation certificate trading)
which will make it more, not less, attractive to invest in energy storage [Armitage and
Biggs (2003)]. The ability of base load conventional plant to sell off peak at less than
marginal cost (€3/kWh) further distorts the economic rationale for integration of
renewables, but supports the argument for storage. This finding, which derives from
point 2 in the summary findings above is illustrated in figure 5.1 (a) and (b) below.
Figure 5.1 : Improving the Attractiveness of Renewable Energy (Supply - Demand Analysis
Case A : Effect of Fixed Feed in Price
Do
Case B: Effect of Energy Storage
So
So
D1
Do
S1
P1
D1
S1
P1
A
A
B
Po
Po
B
C
Qo
Q1
Case A:
• Electricity Demand CurveD0, meets Supply Curve S0 to
give market clearing priceP0 and quantity Q0
Qo
Q1
Case B:
• Electricity Demand CurveD0, meets Supply Curve S0 to
give market clearing priceP0 and quantity Q0
• Regulation in favour of renewables set renewable price at
P1, effectively imposing a tarrif [P0-P1], and causing a
demand shift to D1
• Electricity Demand is not constant at all times, demand
shifts to Demand curve D1 for significant time periods
•Excess quantity of [Q0-Q1] of renewables are produced
•Renewable originated energy, does not wish to participate at
Price P0, but with stored energy particpates at price P1
•Value A accrues as a premium to renewable producers
•Value B is lost by customers
•Quantify corresponding to block C, is not released because
intermittency does not guarantee its availability
•Taxation effect is at least A and B
•Value A accrues as a premium to renewable/ storage
producers, as a supplier surplus
•Quantify corresponding to block B, is displaced as
renewable originated energy now approximates conventional
generating capacity
•Quantity [Q0-Q1] of renewables is produced
•In both cases a demand shift of renewables from Supply
Curve S0 to S1 is desired. Case B is more likely to deliver
this step change
106
Electricity Storage and Wind Energy Intermittency
It should be noted that similar situation can exist under fixed feed in tarrifs, but in this
case the PSO (public service obligation) payment scheme may incent other producers
to pay to store the energy from renewables and then release the energy at times of
higher prices.
Therefore, from a purely microeconomic viewpoint, energy storage has an important
role to play in liberalised electricity markets, where carbon emission constraints are
being imposed. In fact markets where renewable sources are expected to be price
takers, will likely have an added incentive to explore the economic viability of energy
storage technologies and systems.
Potential for renewable energy dispatch
The medium term unpredictability of wind is well documented. However for
significant penetration of wind and other renewable sources (20% and more), a
systematic energy reserve pool is required, to ensure power quality and grid
distribution integrity [Harrison (2003), Garrad Hassan et al (2003) and Nicholson
(2000)].
Very few storage systems can sustain curtailment of conventional generating
equipment over a longer term than say 10-12 hours. Limits apply because of absolute
energy storage capacity and its link to discharge time at rated capacity. Of the storage
systems only hydrogen, or wind hydrogen systems offer the potential of fully
replacing conventional electricity generation, and therefore accruing a ‘capacity
credit’ in their economic analysis [EA Tachnology (1998), Price (2000), DOE (2003)
and Enslin, Knijp et al (2001)]. Economically the renewable-storage energy system
becomes more attractive, as more of the economic benefits associated with zero
marginal cost energy production accrue.
Storage
Typical Output Rating
Typical Storage Rating
Physical
System
(Megawatt)
(Megawatt hours)
Linkage Reqd.
Pumped
100
500- 10000
Yes
Hydro
CAES
25
200
No
Battery
10
100
Yes
Hydrogen
10
Unlimited
No
Table 5.2: Discharge rating and storage capacities for typical systems3
So for instance in a conservative ‘fuel saver’ analysis, the greatest benefit from wind
energy is the reduced fuel costs associated with some reduction in output from
conventional generation plant [Garrad Hassan et al (2003)]. In a more aggressive
analysis, sometimes called a ‘capacity credit’ where wind energy meets all the
dispatch requirements, then fuel saving, externality costs, and capital cost
comparisons become the key drivers of the analysis. The essential requirement is that
hydrogen production and storage can be decoupled, and that large storage facilities
for hydrogen can be constructed.
3
Price (2000)
107
Electricity Storage and Wind Energy Intermittency
If hydrogen systems do attain this potential (<5.1c/kWh integrated cost) then
renewable energy sources can safely replace conventional generation on distributed
and centralised grids [DOE (2003)].
An alternative approach is to develop an interconnected ‘super-grid’ for transmission
of renewable resources on a continental or transcontinental basis. As mentioned
earlier an initiative or its analysis falls outside the terms of reference of this study.
Categorisation of Costs and Prices
For purposes of analysis of economic viability a number of different categories of
costs lie within the model of a storage system. Table 5.3 illustrates the most important
cost categories and a description of their scale:
Costs
Viable Value Comments
Range
Capital Costs
Charging Equipment
<€1000/kW
Scale and technology dependent
(e.g. pumping, or hydrolyser)
Generating Equipment
<€1000/kW
The lowest capital cost plant is typically
CCGT. Depending on scale this can be as
low as €300/kW installed capacity
Scale and technology dependent.
Storage Equipment Cost
Variable Costs
Generating costs variable
Operating Maintenance costs
External Costs
Grid Connection Costs
<€20/kWh
<€0.05/kWh
€0.00/kWh for renewables, €0.02/kWh and
higher for Coal and Gas (cash costs)
<€0.004/kWh Typically decline with scale. (Note that
depreciation is a non cash cost).
Not a direct cost, however many grid
operators may seek to recover costs of
connecting new capacity (particularly in
liberalised market). Both transmission and
distribution grids, are broadly speaking
public goods, and should be designed and
managed to reflect a broad indifference to
the source and destination of electricity.
Fuel Saving Cost
<€0.02/kWh Least benefit accrued from Renewable
generation
Efficiency penalty
<€0.02/kWh Actual penalty for system inability to run
without base generating load at conventional
plant
Carbon Costs
<€0.02/kWh Real costs, but paid by broader group,
associated with burning of fossil fuels. Cash
cost of carbon fuels reflects only, recovery
and distribution costs.
Table 5.3: Overview of most important cost categories for renewable energy4
4
<€50/kW
Price (2000) and Price and Thijssen (1999)
108
Electricity Storage and Wind Energy Intermittency
Many different pricing mechanisms with slightly nuanced meanings are apparent in
different markets. An overview of the most important pricing considerations met in
the both fields of renewable energy pricing and liberalised markets is presented in
Table 5.4
Comments
‘German’ approach to encouraging wind
energy and other renewable production
AER/NFFO tariffs
<€0.10/kWh Contracted prices, similar in effect to fixed
feed in tariffs, where long term price is paid
to producers of wind energy (irrespective of
actual price for electricity in a liberalised
market)
ROC Market price
<€0.05/kWh Traded price for a certificate specifying
production of a single Megawatt Hour of
renewable energy. Distributors have an
obligation to either produce or purchase a
mandated percentage of total electricity as
ROCs
ROC Obligation (Cash in €0.04/kWh
Threshold price below which a ROC can be
price)
cashed in at an issuing authority
Electricity price
<€0.14/kWh Market price of electricity in a liberalised
market, typically power producers bid in
half hourly windows for actual and reserve
power feeds ins. The market price is set
either by pure market or a market clearing
engine (MCE) as in the proposed Irish case.
LMP
<€ TBD
Location Marginal Price (LMP). Price paid
at a node in the network for required power
capacity
PSO tariff
<€ TBD
Difference between renewable (or other
preferred source, such as peat) electricity
price promised to producer and actual price
attainable in a market. This is recharged to
the other producers as a levy which, is
often passed on to the consumer
Charge for reserve
<€ TBD
Causer pays principle will apply.
Table 5.4: Overview of most important pricing categories for renewable energy5
Prices
Fixed feed in tariffs
Value range
<€0.14/kWh
Current Capital costs
Table 5.5 shows capital costs for storage systems most likely to be associated with
wind energy systems. Data is based on best available information as of 2001. There
is a significant scale effect explored in many sources which as expected implies that
the larger the storage system the lower the cost per kW of installed capacity
5
Price (2000), Schoenung (2002), European Commission (2003) and CER (2003d)
109
Electricity Storage and Wind Energy Intermittency
Storage Technology
Capital
Cost
€/kW
wind €1700
Life time limits
Hydrogen
Similar
to
systems
conventional
(electrolyzer/turbine
plant
/storage configuration)
Hydrogen
wind €2400
Similar
to
systems
conventional
(electrolyzer/fuel
plant
cell/storage
configuration)
Pumped hydro
€800 - Similar or better
(pump/storage/turbine) 3000
than
conventional
plants
Compressed air
€600 - Similar or better
(storage/turbine)
€1000
than
conventional
plants
Output range rate
5MW in current production.
No theoretical limit on either
range or output rate
5MW in current production
No theoretical limits
50MWh-15000MWh
total
storage capacity, up to
4000MW/hour highest rates
Up to 1000MWh Limited by
storage
capacity
(underground/some vessels),
rate for hybrid systems up to
400MW/h
Flow batteries (NaS)
€700Up to 50,000 Prototype flow batteries (e.g.
(Combined
1500
cycles at current UK
Regenysys
up
to
charge/storage/
technology
120MWh). Rating up to
discharge)
40MW/h
Lead acid batteries
€300Up to 20,000 Smaller applications installed,
(Combined
charge 1000
cycles at current up to 40MWh with ratings of
/storage/discharge)
technology
10MW/h
Table 5.5: 2001 Capital cost (excluding wind turbine costs) comparison for
appropriate energy storage technologies6
Operating costs and costs per discharge; [4]
Typical operating costs for an integrated energy storage system are very low, and
depend to a large extent on the system integration to demand and supply functions. A
good rule of thumb is to estimate 0.5 c/kWh of output, for a plant of greater than
500MWh per day output, with expected 10 hours per day discharge. Staffing and
maintenance spend are a function of equipment age, type and scale, and therefore
vary within a broad range for different types of storage facility.
Any increase in plant operating time produces a clear scale benefit to the storage
facility. Twelve hours discharge per day is ideal for a coupled system. A hydrogen
based system, may however operate up to 24 hours per day, as the energy source,
storage and regeneration are all decoupled.
6
Schoenung (2002)
110
Electricity Storage and Wind Energy Intermittency
Equipment cost per discharge and depreciation, as noted earlier are not actual cash
costs, so are not included in any analysis. Capital costs are explicitly included in the
viability model, as an initial investment with a specified lifetime and a required
economic rate of return.
Input and output energy costs: [Platts (various) and CER (2003d)]
In considering costs of input energy to a storage system, a clear decision has been
taken to consider only cash costs. Broadly for a system designed to deal with
intermittment supply issues, and where market prices prevail, two cost or price points
present themselves:
• In a curtailment operating environment as envisioned in Ireland, where wind
energy will not be preferentially dispatched the input cost of wind energy is
exactly the marginal cost to produce it, that is €0.000/kWh. This price is
exactly equal to the revenue foregone by diverting the energy to a storage
medium
• In a market environment where renewable wind energy is dispatched or
offered to be bought from a producer at a given price the cost of the input
energy is exactly equal to the LMP (in the Irish case), if the wind producer
turns this down. Alternatively the producer may decide to decline a Fixed
Feed in Tarrif price or an AER price (in the Irish case) if one is in operation.
• In an ROC trading system, stored energy has a cost equal to the cash out cost
of the ROC itself (£30/MWh in the UK). However the ROC is also ‘stored’
and can be resold at market prices.
Similarly output energy costs, are the direct payment that a producer can accrue for
the electricity at the targeted output time. A significant body of research has been
devoted to this topic, with consistent results:
• For a successful peaking strategy, average prices obtained are in the range
€50-60/MWh. This is the maximum output price used as a base point in this
study. Average feed in price data was typically €30-40/MWh
• For a fixed feed in tariff, clearly there is no revenue enhancement (and
therefore little incentive to store renewable energy)
• In some cases (power imbalance) price spikes for electricity are experienced
(although these circumstances like negative prices are actually manifestations
of market failures), allowing very high prices exceeding €100/MWh to be
realised
Cost improvements areas to examine [Baker (2001), Altmann, Niebauer et al
(2000), Nicoletti (1995)]
From the above analysis it can be seen that substantial progress has been made to
reduce the overall cost and improve the financial attractiveness of energy storage.
However it is anticipated that much improvement will accrue over the next years to
further improve the efficiency and cost performance of energy storage. In terms of the
discussion relating to economic underpinnings, this corresponds to an expected
supply side shift from a supply curve S0 to S1 (see Figure 5.1). Topics of current
research are explained in summary below. For each field it should be noted that
111
Electricity Storage and Wind Energy Intermittency
many researchers in academic institutions, government sponsored research and the
private sector are active. The most important areas are:
•
Further reduction in the capital cost of wind energy. With an increase in
potential power per turbine to upwards of 2.5-5MW, and a move to offshore
farming, the likelihood is that the there is potential for a continued decrease in
the cost perMW. It is estimated that the total capital cost of wind energy
production could fall to €800/kW over the next 10 years. Wind energy
marginal production cost will continue to be zero, giving a potential cost of at
least 3 cents per kWh advantage over the most competitive fossil fuel system
(CCGT)
•
Additional storage capacity in wind systems prior to conversion to electrical
energy has not been explicitly modelled in this analysis. However flywheel,
hydraulic pump potential energy storage mechanisms are being developed.
However no reliable cost or operating data on these systems has become
available
•
Improvement in flow battery technology. In the short term it is anticipated that
further electrochemical improvements will occur in the areas of safety, energy
density, and storage lifetime. [private communication and Taylor and
Hoagland (2002)]
•
Both hydrogen wind systems and CAES Hybrids are in their infancy. A
number of investigators have pointed to expected technology, learning curve,
systemisation and scale benefits to be accrued over the next 5 – 10 years
[Nakhamkin et al (2001) and Enslin, Knijp et al (2001)]
•
Development of a systemised approach to pumped hydro distributed storage,
may reduce the overall system cost significantly. Together with more efficient
pumping equipment, and the potential to build plants designed specifically to
combat the intermittency of renewable energy production [Christakis (1996)
and Lyons and Voigt (2003)]
•
Wind forecasting, particularly short term wind forecasting within the context
of market bids for electricity supply is a highly important software enabling
technology [ESB NG (2003a) and Lyons and Voigt (2003)]
•
Development of dynamic response grid transmission and distribution software
is ongoing. The incorporation of renewables is an addition of complexity, and
provides opportunities in the development, maintenance and operation of
appropriate system software [ESB NG (2003a)]
•
Hydrogen production, storage and reuse. The development of a ‘hydrogen
economy’ is particularly important in the context of decoupling the timing of
renewable energy production (wind, solar and biomass) from its use.
Electrolyser technology, hydrogen storage and reintegration via fuel cells,
hydrogen turbines or other mechanisms is a relatively young area of dedicated
112
Electricity Storage and Wind Energy Intermittency
research, but many important advances are already being made. [Liu (2003),
DOE (2003) and Pritchard (2003)]
In summary the expectation is that improvements of the order of 50% in total system
cost, for a combined wind and storage system, will be realised within the optimisation
of the current technology environment.
5.3
Financial model ouputs:
Benefits in the context of a liberalised electricity market
The liberalised market provides a number of opportunities to further advance energy
storage. Specifically the opportunities fall into three interlinked categories [Kroon
(2002)] :
Category 1: Power quality and power management
a) Preventing voltage dip
b) Prevents cascading grid failures
Category 2: Tariff trading:
c) Peak shaving arbitrage
d) Coverage of high value power imbalance requirements
Category 3: Energy storage for the integration of renewable and distributed
generation
e) Improves dispatchability of renewable energy
f) Allows curtailment of base load conventional generation capacity
g) Reduces cost and complexity of integrating renewable energy to transmission,
distribution systems as well as incorporating renewables into the new Market
Arrangements for Electricity
The economic case for energy storage under category 1, power quality management is
well established as the rationale for most older pumped hydro storage systems.
Increasingly CCGT generation is seen as adequate protection against the requirement
for further investment in pumped storage. It should be noted that variable costs for a
CCGT system are high, (June 2003 price of 4c/kWh). Smaller scale pumped hydro
schemes may be more attractive at locational marginal nodes to protect a local or
distributed section of grid. Similarly explicitly utilising zero marginal cost, wind
energy as the charging power for an existing large scale storage system, significantly
improves its inherent profitability.
The economic case for tariff trading (category 2) is readily established. Using some
historic data from a client example the arbitrage value is demonstrated to be up to
€20/MWh, (input energy cost – output energy price) The net benefit means that a
peak shaving battery system is viable, even when linked to conventional generation
technology. The key variables affecting the viability of a grid input and output
storage system are the number of charges and discharges per day. Prior to the project
closure, it was anticipated that the operating data from the Regenysys system will
113
Electricity Storage and Wind Energy Intermittency
give some insight into these essentially market driven economics. [Kroon (2002),
Pritchard (2003)]
Compressed Air Energy Storage can also be used in conjunction with gas turbines to
provide power management and longer term power quality security. A number of
systems have been developed to utilise the compressed air to improve gas turbine
efficiency to upwards of 70%. The fundamental trade off is related to cost of storage.
It should be noted however that CAES systems are being built using pressure vessels
as well as large underground storage facilities [Nakhamkin et al. (2001)]
Large scale energy storage for renewable applications, where the ultimate goal is to
integrate renewables into the conventional grid and cause curtailment of conventional
generation capacity, requires significant storage capacity. In reality the only systems
which are suitable are those where the energy production and utilisation are
substantially separated. Hydrogen wind systems are the most likely means to achieve
this goal of conventional generation curtailment. The key cost determinants in the
valuation models are:
1. Cost per MW for storage charging
2. Capital cost of equipment €/kWh
3. Cost per MWh /day for storage
4. Round trip efficiency
However on an ongoing basis, both pumped hydro and compressed air, offer the
potential to significantly increase the reliability and thus the dispatchability of
renewable energy. Both are limited only to the extent of the available storage
reservoir. In the case of pumped hydro, smaller scale systems can be linked to
individual wind farm sites, and may allow for much higher storage to discharge ratios
required to ensure power availability.
Wind hydrogen system
The economic model, developed during this study, produces some important results in
terms of the viability of the systems. The variables used and associated ranges are
tabulated in Appendix 2. Assuming that it is technically feasible and meets the basic
requirements to allow the electricity to be dispatched on equal terms to other power
producers, the viability of a storage system, depends quite heavily on four factors.
1. The price enhancement available between variable charge cost (foregone
price) and discharge attained price
2. The average time period when the system can discharge to an enhanced
revenue environment
3. The capital cost of the facility (charging equipment, storage equipment, and
discharge equipment)
4. The efficiency of the system
In this context hydrogen wind systems are potentially viable. Because of the
independence of storage from charge and discharge, these systems are the most
suitable (together with battery storage) for peak shaving. However given sufficient
storage, wind hydrogen can be as flexible as a methane driven CCGT. For all systems
the expected realisable price enhancement is modelled as either at €30 or €40 per
114
Electricity Storage and Wind Energy Intermittency
MWh, depending on the time of discharge[Nakhamkin et al (2001), Baker (2001),
Platts (various), Lyons and Voigt (2003), Crudden (2000), Taylor and Hoagland
(2002), Sandia Lab (1994)].
The economic evaluation of wind hydrogen used a number of static and variable
inputs. However the key drivers of viability are system cost, and system efficiency.
Total system cost is expected to reduce significantly but current costs for an
integrated production, storage, turbine/engine system are close to €2,200 per kW.
Potential savings have been evaluated by a number of researchers, and these
expectations have been factored into ongoing business expectations. A combined
electrolyser, storage system total cost of less than €1,500 per kW may be reached in
the next three years. [Pritchard (2003), CER (2003b)]
Figure 5.2 illustrates the output from the economic model as a viability surface, where
NPV of a hydrogen wind system is given in terms of variance in system capital cost
and system efficiency. These two variables are the most easily influence by a
concerted drive to develop this technology. Typically with capital cost at or around
€1,500/kWh and system efficiency at approximately 40%, the system becomes
profitable in current market pricing conditions.
Economic Model of Hydrogen Wind System
600
800
1,000
1,200
1,500
System Capital Cost €/kwh
1,700
0.75
400
0.55
6.0
4.0
2.0
0.0
2.0
4.0
6.0
8.0
10.0
0.35
NPV € Millions
System
Efficiency
Notes: 1. Key static inputs: 5 MW Electrolyser, with ICE Turbine Gensets, or Fuel Cells, Advanced storage medium, or full buffer
2. Input cost of wind power €0.0/MW (marginal cost), output price €40/Mwh (Nogales)
3. ICE engine uses electrolyser output at ambient pressure, output diverted for 10 hours per day (40MW/day)
4. Equipment cost from (Pritchard, Liu and proc. EERE)
5. Financing cost 7% pa., lifetime 20 years
Figure 5.2 Economic model of wind hydrogen system
The economic viability increases for these systems if variance is allowed on hours of
daily production. Because of decoupling the energy generation segment of a
Hydrogen Wind system, the energy production or availability to a grid begins to
resemble availability for a conventional generation facility. In fact many pumped
hydro storage systems (e.g. Turlough Hill) now operate up to 16 hours a day in
discharge mode. Table 5.6 illustrates the effect by way of three case examples.
115
Electricity Storage and Wind Energy Intermittency
Capital cost for a hydrogen turbine or hydrogen engine system will likely approach
total cost for a methane (natural gas) turbine of approx. €400/kW. Future turbine
systems will generate significantly less NOx pollutant than current generation
turbines.
Capital cost
Price enhancement
System efficiency
Operating
hours/day
NPV
Current case
(intermittency
arbitrage)
Near term case
(intermittency
arbitrage)
€2500 /kW
€0.02/kWh
30%
10
€1800/kW
€0.03/kWh
36%
10
Medium term case
Decoupled storage
(fully dispatchable
plant)
€1400/kW
€0.03/kWh
47%
20
(€11.2m)
(€6.9m)
€1.1m
Note: 5MW matched hydrolyser and generation equipment, storage estimated at €2/kWh capacity
Table 5.6: Large storage, long discharge hydrogen wind systems
Fuel cells are an alternative to a gas turbine, and offer the prospect of significantly
higher efficiencies. However fuel cell prices are currently prohibitively expensive and
not commercially available. Further this technology is still in intense development, by
companies such as Ballard Power and Plug Power.
Electrolyser efficiencies of higher than 70% seem to be attainable from a number of
manufacturers. However the production of hydrogen from wind energy is a relatively
new market for suppliers, who have previously concentrated on hydrogen purity as
the critical performance indicator of their products and systems.
As can be seen, within current constraints the complete system struggles to be viable.
However, with small incremental increases in efficiency and reductions in capital
cost, both of which are expected by many experts the system becomes economically
viable at relatively low hydrogen production of 8 hours per day from a connected
wind farm site.
The viability becomes more compelling if a larger facility is constructed (Large Scale
Wind Hyrdrogen – LSWH) with more utilisation of the expensive capital plant (33%
in the analysis presented here, but potentially as high as 80% to 90%)
Wind pumped hydro system
Pumped hydro facilities have typically been constructed with a view to ensuring
power quality on a transient basis in a large integrated grid. In the last ten years
however, CCGT spinning reserve has become more viable and the smaller scale of
incremental capacity requirements for a conventional grid can be well met from
CCGT, or in extremis, oil fired generation plant.
Along with capital costs and efficiencies, the ratio of storage capacity to generating
capacity is central to improving the dispatchability of renewable energy. Storage
capacity equating to two days of rated reserve output from a plant, should be adequate
116
Electricity Storage and Wind Energy Intermittency
for the wind – hydro power integrated system to participate in a liberalised electricity
market on an equal dispatch paradigm. For a 10MW output plant this implies a
storage capacity of 480MWhs. Interestingly storage capital costs for both pumped
hydro and flow battery are expected to be higher than for hydrogen systems. This
does not take into account the operational storage costs which will tend to be higher
for hydrogen and battery technologies than for pumped hydro (where these costs
approach zero).
Economic Model of Small Scale Pumped Hydro
NPV € Millions
8.0
6.0
4.0
2.0
0.0
2.0
4.0
6.0
8.0
10.0
400
600
800
1,000
System Capital cost €/kwh
1,200
1,500
1,700
1,900
0.35
0.55
0.75
System
Efficiency
Notes: 1. Size data : 11 MW rated facility
2. Assumed output at 10 hours per day
3. Attained price enhancement of €30 per megawatt
4. Financing cost 7% pa., lifetime 20 years
Figure 5.3 Economic model of wind pumped hydro system
Similar models have been constructed for other storage technologies, however the
operating data on charge and discharge times for flow batteries are not sufficiently
disseminated to allow for publication of this data.
Compressed air storage relies on the availability of suitable storage facilities.
However some installations are currently using pressure vessels to store compressed
air. The central attraction of a compressed air system is that it allows a conventional
CCGT to work at higher efficiencies (up to 80%). In measurement terms, a CAES
gives an reduction in cost per kWh of approximately 2cents. Running continuously a
50MW turbine will save approximately €8.4m in running cost per year, if
continuously fed with compressed air.
External benefits
The external benefits of energy storage have not been explicitly modelled in the
financial viability study, however there are a number of important external benefits
117
Electricity Storage and Wind Energy Intermittency
from the integration and maturation of an energy storage capability in the Irish
context.
•
An integrated storage strategy improves the reliability of wind energy and
removes many, and potentially all, issues around intermittency of wind energy
as a primary source of energy. This then allows wind to be viewed in dispatch
terms, as equivalent to conventional energy supply, thus providing the basis
for higher investment and more attractive returns from wind energy.
•
By allowing wind energy to be integrated to the overall energy requirement,
energy storage can reduce the need for additional conventional generation in
the medium and long term. Similarly given the characterisation of Ireland as
the ‘Saudia Arabia of Wind’, there is potential for Ireland to become a net
exporter of renewable energy
•
Energy storage by increasing wind energy installations, allows Ireland to
reach its EU and Kyoto commitments for greenhouse gas emission reduction,
on time and with reduced cost. These commitments are in themselves proxies
for the actual external costs attributed to fossil fuel usage.
•
Development of an intellectual knowledge base in a new industry such as
energy storage is valuable for Ireland, and its migration to a knowledge
economy
•
Energy storage linked to wind energy enables distributed transmission and
distribution systems for electrical energy. At one level this will reduce the
cost associated with distribution to remote areas, and at another level it will
reduce the cost associated with provision of back up generating capacity
(often diesel) at non grid connected sites.
118
Electricity Storage and Wind Energy Intermittency
Chapter 6 STRATEGY
The short to medium term strategy focuses on the utilisation of mature electricity
storage technologies where performance characteristics and costs are better known
and more clearly understood. The longer term strategy concentrates on technologies
that are not yet mature but are potentially more promising in terms of their suitability
in addressing wind energy intermittency.
The strategies focus on the storage technologies themselves and how they will operate
within the context of electricity network and electricity market developments.
Short to medium term strategy
The key elements of the short term strategy are :1. Pumped hydro resource study. A significant theoretical resource (up to
1,000MW) has been identified within the context of this study. A study is
required to determine the practicable pumped storage potential, taking into
account technical and non-technical constraints. It is important to distinguish
in this study between a) the potential for pumped hydro storage plants
associated with individual wind farms, and b) the potential for pumped hydro
plants that provide storage capability for a group of wind farms (data from
Turlough Hill could be utilised as a reference case in this analysis). In the case
of the latter, the storage plant will address the issues of wind energy
intermittency associated with load levelling and back-up requirements.
whereas the former has also the potential of addressing local network
constraints to wind energy integration.
2. Compressed air energy storage. The potential for compressed air energy
storage should be undertaken providing details of optimum locations close to
gas generators with underground reservoirs. This will entail geological
surveying and electromechanical modifications to existing or proposed gas
fired generators. It may be beneficial draw on the results of Optimum
Energy’s work if the publication timeframe is appropriate. The requirement to
carry out a CAES feasibility study should also be considered by CER in the
licensing process for new gas fired electricity generators.
3. System modelling. The use of storage needs to be considered in the context of
an integrated approach to dealing with wind energy interactions with the
electricity network. This will require the development of real time energy
systems models linked to pending grid modelling studies and incorporating the
use of wind energy forecasting. It should also consider the use of methods for
addressing intermittency other than storage (for example open cycle gas or and
East West interconnector), that fell outside of the scope of this study;
4. Grid upgrading programme. The current extensive grid upgrading programme
currently underway should be reviewed to take account of the prolific increase
and concentration in anticipated future wind energy production. This
programme review is a key recommendation of the Renewable Energy
Strategy Group and is separate to the agreed mechanism addressing the
challenge that existed for developers where they must raise the entire capital
119
Electricity Storage and Wind Energy Intermittency
expenditure for any upgrade forming part of a potentially shared connection
with money subsequently remitted as others connect to the facility.
5. Demonstration Projects. The purpose of these projects is to link mature
storage technologies with wind energy to demonstrate the technical and
economic viability of the complete system. This will drive the learning curve,
reducing capital costs and increasing future operational efficiencies. The
projects should only proceed following a detailed, fully costed technical and
financial feasibility study
a. Wind + Small scale pumped hydro
This demonstration project should provide insights into the potential of
small scale pumped hydro to address wind energy intermittency in Ireland.
The pumped hydro plant should be situated in close proximity to the wind
farm in order to assess the ability to provide reserve in addition to
providing back-up for and storing the wind energy. The provision of
detailed technical information should be a condition of financial support
associated with the project. The funding support provided should be linked
specifically to the pumped hydro storage elements of the system.
b. Wind + Compressed air energy storage
This demonstration project should provide insights into the potential of
compressed air energy storage to address wind energy intermittency in
Ireland. The CAES should be situated in close proximity to a gas fired
electricity generation plant, and where possible, also close to the wind
farm. The provision of detailed technical information should be a
condition of the project. The funding support provided should be linked
specifically to the CAES elements of the system.
Long term strategy
The key elements of the long term strategy are
1. Linking wind energy storage and the hydrogen economy. A study will be
required to detail the synergies between hydrogen production in the context of
wind energy storage and the development of the hydrogen economy. In
particular the anticipated advances in hydrogen fuel cell technologies will
increase the value of hydrogen and as a result improve the economics of wind
hydrogen systems. It is crucial to investigate beyond the hype surrounding the
hydrogen economy and use a sound basis for this work. It will only be
meaningful to carry out this study when more data becomes available from
various detailed studies on the future of hydrogen and fuel cells that are
currently underway.
2. Demonstration projects. The purpose of these projects is to assess more
flexible storage technologies that have the potential to address wind energy
intermittency more completely by dealing short term fluctuations as well as
providing load levelling and back-up. These technologies are not yet
technically or economically proven in Ireland. It is recommended that research
as well as demonstration projects be considered in this category, given the
stage of development of these technologies. In addition, it is recommended
that the results of EU and extra-EU projects be disseminated in Ireland as they
become available.
120
Electricity Storage and Wind Energy Intermittency
The projects should only proceed following a detailed, fully costed technical
and financial feasibility study and as in the case of the other demonstration
projects proposed, the provision of detailed technical information should be a
condition of the project
a. Wind + flow battery
Despite the withdrawal of Regenesys, flow batteries are still a strong
contender as a potential storage solution to wind energy intermittency,
with both VRBs and ZnBr batteries targeting wind energy projects. It is
recommended that the choice of flow battery type should be made based
on project proposals submitted, rather than being prescribed in advance.
b. Wind + hydrogen engine
While engines are less efficient than fuel cells, a wind hydrogen engine
demonstration project can be seen as a transition stage to a wind hydrogen
fuel cell project and can provide some valuable information on the stage of
progress in electrolyser and engine technology, as distinct from fuel cell
technology.
c. Wind + hydrogen fuel cell
The absence of detailed information from working wind hydrogen fuel cell
systems is a key gap in this area. There is a lot of international focus that
should reduce the costs of these systems. A demonstration project would
provide clarity relating to the performance of the individual components,
overall system efficiency, current costs and valuable insights into what
may be required to make these systems viable.
In summary, the energy storage sector is central to the full integration of wind energy
generation. There are a number of appropriate technologies, the most attractive of
which from a flexibility viewpoint is wind hydrogen, because of the ability to
decouple the input power, output power and storage capacity. Furthermore, wind
hydrogen systems are attractive from the standpoint of achieving zero emissions
energy. It has not yet matured from an economic perspective however and the overall
energy efficiency remains poor.
Pumped hydro systems and compressed air systems have the advantage of technical
maturity, economic viability and operational experience and are therefore viewed as a
realistic and appropriate first stage in the development of an energy storage solution
to wind energy intermittency.
121
Electricity Storage and Wind Energy Intermittency
122
Electricity Storage and Wind Energy Intermittency
REFERENCES
Abaoud, H. Steeb, H. (1998) The German-Saudi Hysolar Project. International
Journal of Hydrogen Energy, Vol. 22, No 6, pp. 445-449, 1998.
Agbossou, K. Hamelin, J. Chahine, R. (2000) Fuel Cell and Hydrogen Storage for
Stand Alone Renewable Energy System. International Journal of Renewable Energy
Engineering, Vol. 2, No. 3, 2000.
Altmann, Niebauer et. al. (2000) Wind Hydrogen Supply of Electricity Markets –
Technology –Economics, Proc. Wind Power 21st Century, September 2000
Armitage & Biggs (2003) The UK Renewables Obligation: What’s holding back the
ROC market in the UK power sector?, Refocus Mar 2003
Akhil, A. Swaminathan, S. Sen, R. (1997) Cost Analysis of Energy Storage Systems
for Electric Utility Applications. Sandia Report, SAND97-0443 l UC-I 350, 1997.
Akhil, A., Kraft, S. 1997b) Battery Energy Storage Market Feasibility Study. Sandia
Report, SAND97-1275/1 UC-I 350, 1997b.
Baker, J, EA Technology, (2001) Analysing the future for electricity storage
technologies, IIR conference on Commercially Viable Electricity Storage, January
2001
Barthels, H. Brocke, W. Bonhoff, K. (1996) An Autonomous Energy Supply System
Comprising Photovoltaics, Electrolytic Hydrogen, Fuel Cell. In the 11th World
Hydrogen Energy Conference, Stuttgart, 1996.
Baxter, R. Makansi, J. (2002) Energy Storage. The Sixth Dimension of the Electricity
Value Chain. Pearl Street Inc. 2002.
Blackaby, N. (2002) VRB Technology Comes to the Fore. Power Engineering
International May, 2002.
Boyes, J. (2000) Energy Storage Systems Program Report for FY99. Sandia Report,
SAND2000-1317, 2000.
Breeze. P. (1998) Power Generating Technologies – Evaluating the Cost of
Electricity. FT Energy Report. ISBN 1 84083 038 7, 1998
Buckles, W. Hssenzahl, W. (2000) Superconducing Magnetic Energy Storage. IEEE
Power Engineering Review, Vl. 20, N.5, pp. 216-220. 2000.
Butler, P. Miller, J. Taylor, P. (2002) Energy Storage Opportunities Analysis. Phase
II Final Report: A Study for the DOE Energy Storage Systems Program. Sandia
Report SAND2002-1314. Unlimited Release. May 2002.
Carnegie Mellon University (2001) Transition to a Hydrogen Based Economy: The
Next Ten Years 2001
CER (1999) Proposals for a Transitional Electricity Trading System for Ireland,
CER/99/02, September 1999
CER (2003) Funding of Grid Upgrade Development Programme for Renewables –
Decision Paper available from http://www.cer.ie/CERDocs/cer03090.pdf
CER (2003a) Quarterly Bulletin, Quarter 2, 2003 available from
http://www.cer.ie/cerdocs/cer03157.pdf
123
Electricity Storage and Wind Energy Intermittency
CER (2003b) Implementation of the Market Arrangements for Electricity (MAE) in
relation to Renewables, CHP and Distribution-connected Generation - An MAE
Consultation by the Commission for Energy Regulation Under S.I. 304 of 2003
available from http://www.cer.ie/cerdocs/cer03253.pdf
CER (2003c), ESBNG Proposal to Limit New Wind Connections, CER response to
ESBNG, December 2003
CER (2003d), Market Arrangements for Electricity in Ireland (MAE), September
2003
CER (2003e) MAE Detailed Design – Response to Consultation, 23 October 2003
CER (2004) Wind Generator Connection Policy – Draft Direction by the Commission
for Energy Regulation available from http://www.cer.ie/cerdocs/cer04183.pdf
Christakis, D. Psofogiannakis, M. (1996) Reduction of Electricity Cost by
Implementing Pump-Storage Units in Crete Island. Proceedings of U.P.E Conference,
Heraklion Crete, 1996.
Christakis, D. Fassoulas, V. Sifakakis, K. (1997) Wind Energy Conversion Systems
Combined with Pumped Storage Systems (PSS) for Small Isolated Power Production
System. The European Congress on Renewable Energy Implementation, 5-7 May,
Athens, 1997.
Christakis, D. Minadakis, J. Nikiforakis, M. Fasoulas, M. (2001) Towards 100% RES
Supply for the Electrification of Crete. International Conference Renewable Energies
for Islands, Chania – Crete, Greece, 14-16 June, 2001.
Collinson, A. Stones, J. Morgan, T. (1999) Electrical Energy Storage (Annex IX
Phase 2) Final Report. EA Technologies, Report No. 5073, Project No. 42050,
November 1999.
Collinson, A. (2000) The Costs and Benefits of Electrical Energy Storage.
Proceedings of the Electrical Energy Storage Systems Applications and Technologies
International Conference 2000 (EESAT2000), Orlando, September 18-20 2000.
Cropper, M. (2001) Solid Oxide Fuel Cell Technology in Europe: Trends and
Developments. Fuel Cell Today, October 2001.
Cruden (2000) Opportunities for Energy Storage Devices Operating with Renewable
Energy Systems, Center For Economic Renewable Power Delivery, EESAT 2000
Department of Communications, Marine and Natural Resources (2002) Grid Upgrade
Development Programme – Steering Group Report available from
http://www.dcmnr.ie/files/Steering%20Group%20final%20report%20(2003).doc
Department of Communications, Marine and Natural Resources (2003) Options for
future renewable energy policy, targets and programmes.
Department of the Environment and Local Government (2000) National Climate
Change Strategy available from
http://www.environ.ie/DOEI/DOEIPol.nsf/wvNavView/Climate+Change?OpenDocu
ment&Lang
Dept of Taoiseach (2003) New Connections – Second Progress Report on the
Information Society, 2003
DOE (2003) Procedings of Workshop on Wind Hydrogen Systems, September 2003
124
Electricity Storage and Wind Energy Intermittency
Dutton, A. et al. (1996) Hydrogen Generation from Stand-Alone Wind-Powered
Electrolysis Systems. Final Report, JOU2-CT93-0413. 1996.
EA Technology (1998) Electrical Energy Storage Systems and Technologies, Chester
June 1998
EA Technology (1999) The costs and benefits of electrical energy storage, I.Mech.E
Seminar, 1999
Enslin, Knijp et al. (2001) Commercially viable electricity storage, Paper presented at
Transmission and Distribution Power, January 2001
Enslin Johan (2004) In store for the future? Interconnection and energy storage for
offshore wind farms. Renewable Energy World Vol 7 No. 1 pp 104 – 114. James and
James.
ESB (2003) Strategy to Reduce Emissions of Transboundary Air Pollution by 2010
available from
http://www.environ.ie/DOEI/DOEIPol.nsf/0/25c623e11004192380256b53005d1866/
$FILE/ESB.doc
ESB National Grid (2003) Generation Adequacy Report 2004 – 2010
ESB National Grid (2003a) Interim policy on wind farm connections
ESB National Grid (2004) Wind farm power station grid code provisions
European Commission (2003) Non nuclear energy research topics, November 2003
European Commission (2003a) TRECKIN project, Creating demand for renewable
energy certificates, EU Commission, September 2003
European Commission (2002) The future potential of energy storage. Published by
DG Research, available from
(http://europa.eu.int/comm/research/energy/nn/nn_rt_st2_en.htm)
Friedland, R. Speranza, A. (1999) Integrated Renewable Hydrogen Utility System.
Proceedings of the 1999 DOE Hydrogen Program Review, 1999.
Galli, S. Stefanoni, M. (1997) Development of a Solar-Hydrogen Cycle in Italy.
International Journal of Hydrogen Energy, Vol. 22, No 5, pp. 453-458, 1997.
Gandy, S. (2000) A Guide to the Range and Suitability of Electrical Energy Storage
Systems for Various Applications, and an Assessment of Possible Policy Effects.
Imperial College of Science, Technology and Medicine, London, 2000. Master
Thesis.
Garrad Hassan, ESBI, University College Cork. (2003) The Impacts Of Increased
Levels of Wind Penetration on the Electricity Systems of the Republic of Ireland and
Northern Ireland. Final Report Document 3096/GR/04, 2003.
Gonzalez A. and Ó Gallachóir B.P. (2003) Fuel Cells – Current Status presentation to
Institution of Engineers of Ireland April 7th 2003, Dublin.
Gonzalez, A., McKeogh E., and Ó Gallachóir B. (2003), The role of hydrogen in high
wind energy penetration electricity systems: The Irish case Renewable Energy 29 471
- 489
125
Electricity Storage and Wind Energy Intermittency
Gordon, S. Falcone, P. (1995) The Emerging Roles of Energy Storage in a
Competitive Power Market: Summary of a DOE Workshop. Sandia Report. SAND958247, 1995.
Gyuk, I. (2002) The Power of Energy Storage. Power Quality, March 2002.
Harrison, L (2003) European Hydrogen Vision Under Fire, Windpower Monthly,
May 2003
Headifen R. (1994) Flywheel energy storage for wind turbines. Proceedings of the
29th Intersociety Energy Conversion Engineering Conference. Vol. 4. 1994. pp. 13741379, 1994.
Herr, M. (2002) Economics of Integrated Renewables and Hydrogen Storage Systems
in Distributed Generation. Imperial College of Science, Technology and Medicine,
University of London, June 2002.
Howley, M. & Ó Gallachóir, B. (2002), Energy in Ireland – Trends, issues and
indicators 1990 – 2001, Dublin: Sustainable Energy Ireland available from
http://www.sei.ie/uploads/documents/upload/publications/AR2002.pdf
International Energy Agency (2003), Energy Policies of IEA Countries – Ireland 2003
Review
International Energy Agency (2003a). Implementing Agreement on Energy
Conservation through Energy Storage. Annual Report 2002. OECD/IEA Committee
on Energy Research and Technology End-Use Working Party, April 2003.
Ingram, A. (2000) What a "Good" Energy Storage System Would Look Like to a
Large Utility. Proceedings of the Electrical Energy Storage Systems Applications and
Technologies International Conference 2000 (EESAT2000), Orlando, September 1820, 2000.
Krom, L. (1998) Renewable Hydrogen for Transportation Study, Final Report, L&S
Associates, 1998.
Kroon, Thijssen (2002) Commercial Viability of Flow Batteries, ESSAT 2002
Linden, D. Reddy, T. (2001) Handbook of Batteries. Third edition. New York.
McGraw-Hill. 2001. ISBN 0-07-135978-8
Liu. E, (2003) GE Global Research, Large Scale Wind Hydrogen Systems, DOE
Workshop, September 2003
Lotspeich, C. (2002) A Comparative Assessment of Flow Battery Technologies.
Proceedings of the Electrical Energy Storage Systems Applications and Technologies
International Conference 2002 (EESAT2002), San Francisco, April 15-17 2002.
Lyons, Voigt, GE Wind Engineering Research Team, December 2003
Madlener R, Stagl S., (2003) Promoting Renewable Electricity Generation through
Guaranteed Feed in Tarrifs vs. Tradable Certificates, Proc. 3rd. Conference for Soc.
for Eco. Economics, May 2003
Manwell J. F, McGowan J. G. & Rogers A. L. (2003) Wind Energy Explained
published by Wiley
126
Electricity Storage and Wind Energy Intermittency
Menictas C. Hong D. Yan Z. Wilson J. Kazacos M. Skyllas-Kazacos M. (1994) Status
of the Vanadium Redox Battery Development Program. In Proceedings of Electrical
Engineering Congress '94, pp. 299–306, Sydney, Australia, 1994.
Menzl, F, (2003) A zero-emmission autarchic energy supply for communities in rural
areas, Proc. World Wind Energy Conference, Cape Town, November 2003
Meurer, C. Barthels, H. Brocke, W. Emonts, B. Groehn, H. (1999) Phoebus: An
Autonomous Supply System with Renewable Energy. Six Years of Operational
Experience and Advanced Concepts. Institute for Materials and Processes in Energy
Systems, Research Centre Julich, Germany, 1999.
Meurer, H. (2000) Auslegung Saisonaler Wasserstoffspeicher fur die
Enerigieversorgung aus Solartrahlung, Windkraft und Umgebungswarme. Institute
fur Werkstoffe und Verfahren der Energietechnik, Forschungszentrum Julich, 2000.
Nakhamkin et al., Hybrid Plant Technology for Distributed Power Generation, ESPC
Inc.
Nakhamkin, M. Wolk, R. (1999) Compressed Air Inflates, Gas Turbine Outputs.
Power Engineering, pp. 38-40, March 1999
Nicholson G, (2000) Penetration of Wind Energy in Ireland, February 2000
Nicoletti, (1995) The Hydrogen Option for Energy: A Review of Technical,
Environmental and Economic Aspects. Intl. Journal of Hydrogen Energy 1995, 20
(10):759-765
Oak Ridge National Laboratory (ORNL). (1997) Applying Wind Turbines and Battery
Storage to Defer Orcas Poser and Light Company Distribution Circuit Upgrades.
ORNL-Sub/96-SV115/1, 1997.
PA consultants (2003) The operation of an electricity market, CER conference 23
October 2003
Padro, C., Putsche, V. (1999) Survey of the Economics of Hydrogen Technologies.
National Renewable Energy Laboratory, Report NREL/TP-570-27079, 1999.
Platts (various), European Power Daily.
Powercorp (2001) The Denham Wind/Diesel System, January 2001
Prescher, K. Nick, W. Vollmar, H. Kleinmaier, M. Radtke, U. (1995) Supraleitung in
der Energietechnik II, VDI-Verlag: 115, 1995.
Price, A. Thijassen, G. Sysmons, P. (2000) Electricity Storage, a Solution in Network
Operation? Paper presented at Distributech Europe, October 2000.
Price, A. (2000) Recent Developments in the Design and Applications of a UtilityScale Energy Storage Plant. Proceedings of Proceedings of the Electrical Energy
Storage Systems Applications and Technologies International Conference 2000
(EESAT2000), Orlando, September 18-20.
Pritchard D, (2003) Wind Hydrogen Hunterston Project, DOE Workshop, September
2003
Rambach, G. (1999) Integrated Renewable Hydrogen Utility System. Proceedings of
the 1999 DOE Hydrogen Program Review, 1999.
127
Electricity Storage and Wind Energy Intermittency
Ramesh M, (1999) Renewable Energy: Learning from the British experience,
Business Line, November 1999
Renewable Energy Information Office (1999) Working Group on Grid Connection
Issues Relating to Renewable Energies – Final Report
Renewable Energy Strategy Group (2000) Strategy for Intensifying Wind Energy
Deployment available from http://www.dcmnr.ie/display.asp?pg=800
Renewable Energy World (2003) Another Record Year for European Wind,
November 2003
Sandia National Laboratories (1994) The Emerging Roles of Energy Storage in a
Competitive Power Market, Sandia Lab, DOE 1994
Sandia National Laboratories (2002), Energy Storage Opportunity Analsysis
SAND2002-1314s
Sandia National Laboratories and Black & Veatch. (2001) Boulder City Battery
Energy Storage Feasibility Study. Project Number 66761.0010, Final Report. 2001.
Scherer, G. Newson, E. Wokaun, A. (1999) Economic Analysis of the Seasonal
Storage of Electricity with Liquid Organic Hydrides. International Journal of
Hydrogen Energy, Vol. 24, pp. 1157-1169, 1999.
Schoenung, S. (2001) Characteristics and Technologies for Long vs. Short-Term
Energy Storage. Sandia National Laboratories. Report SAND2001-0765, March 2001.
Schoenung, S. (2002) Characteristics of Energy Storage Technologies for Short- and
Long-Duration Applications. Proceedings of the Electrical Energy Storage Systems
Applications and Technologies International Conference 2002 (EESAT2002), San
Francisco, April 15-17 2002.
Schucan, T. (2000) Case Studies of Integrated Hydrogen Energy Systems.
International Implementing Agreement, 2000.
Sels, T. Dragu, C. Van Craenenbroeck, T. Belmans, R. (2001) Electrical Energy
Storage Systems: Existing Systems versus Newest Systems - an Overview.
International Conference Power Generation and Sustainable Development (AIM),
Liège, Belgium, October 8-9, 2001, pp. 215-220.
Semadeni, M. (2003) Energy Storage as an Essential Part of Sustainable Energy
System. A Review on Applied Energy Storage Technologies. CEPE Working Paper
No. 24, Zurich, May 2003.
Streit and Wang (2000) Advanced Energy Storage – Focus on Electrochemical
Engine, US Department of Energy, July 2000
Swaminathan, S. Sen, R. (1997) Electric Utility Applications of Hydrogen Energy
Storage Systems. US Department of Energy. DOE/GO/10/70-T19, 1997.
Szyszka, A. (1998) Ten Years of Solar Hydrogen Demonstration Project at Neunburg
Vorm Wald, Germany. International Journal of Hydrogen Energy, Vol. 23, No 10, pp.
849-860, 1998.
Tanaka, T. Kurihara, I. (1998) Market Potential of Utility-Purpose Energy Storage in
Japan up to the Year 2050. Proceedings of the Electrical Energy Storage Systems
Applications and Technologies International Conference 1998 (EESAT98), Chester,
UK, June 16-18, 1998.
128
Electricity Storage and Wind Energy Intermittency
Taylor, P. Johnson, L. Reichart, K. DiPietro, P. Philip, P. Butler, J. (1999) A Summary
of the State of the Art in SMES, FES and CAES Systems. Sandia National Laboratories
Report SAND99-1854 July 1999.
Taylor and Hoagland (2002) Using energy storage devices with wind energy for
arbitrage, EESAT2002
Taylor, R. Energy Storage For Ancillary Services. (2002) Proceedings of the
Electrical Energy Storage Systems Applications and Technologies International
Conference 2002 (EESAT2002), San Francisco, April 15-17 2002.
Tennessee Valley Authority (TVA). (2001) Environmental Assessment: The
Regenesys™
Energy
Storage
System,
August
2001.
www.tva.gov/environment/reports/regenesys.
Tennessee Valley Authority (TVA). (2002) Environmental Assessment 20-MW
Windfarm and Associated Energy Storage Facility. April 2002.
Ter-Gazarian, A. (1994) Energy Storage for Power Systems. Peter Peregrinus Ltd,
1994.
Theodoropoulos, P. Mantas, Z. Zervos, A. (2003) Integrated Power System of Serifos
Island with High RES Penetration Using Pump-Storage. International Conference
Renewable Energy Sources for Islands, Tourism and Water Desalination. Crete,
Greece, 26-28 May 2003.
Van der Linden, S. (2002) CAES for Today’s Market. Proceedings of the Electrical
Energy Storage Systems Applications and Technologies International Conference
2002 (EESAT2002), San Francisco, April 15-17 2002.
Wagner, R. Schroeder, M. Stephanblome, T. Handschin. E. (1999) A Multifunctional
Energy-Storage System with High-Power Lead–Acid Batteries. Journal of Power
Sources, 78, pp 156–163, 1999.
Yurek, G. Howe, J. Kehrli, B. (1999) Recent Advances in Superconductivity for
Electric Utility T&D Systems: Laying the Foundation for the "Electricity
Superhighway". Edison Electric Institute's Transmission and Distribution Spring
Conference, 1999.
Zaininger, H. Barnes, P. (1997) Applying Wind Turbines and Battery Storage to Defer
Orcas Power and Light Company Distribution Circuit Upgrades. Oak Ridge National
Laboratories, Oak Ridge, TN: March 1997. Report ORNL-Sub/96-SV115/1.
Zink, J. (1997) Who Says You Can't Store Electricity? Power Engineering, March
1997, pp. 21-25.
129
Electricity Storage and Wind Energy Intermittency
130
Electricity Storage and Wind Energy Intermittency
APPENDIX 1. WIND FARMS WITH PLANNING PERMISSION
Table A.1 Location of wind farms with planning permission
Capacity
(MW)
Turbines
Under
Appeal?
Cape Clear Island
Cork
Dromourneen, Deereennacno, Glanaphuca Cork
Bere Island
Cork
Coomatallin
Cork
Lahanaght, Derryclough
Cork
Coomatalin, Kippagh
Cork
Garranure, Kilvinane
Cork
Cloghmacsimon
Cork
Cappaboy/Curraglass/Maugha
Cork
Milleeny
Cork
Inchamore
Cork
Cahernafulla, Kilberrihert
Cork
Knockraheen, Carriganimmy
Cork
Pluckanes West
Cork
Gneeves
Cork
Gneeves
Cork
Scartbarry
Cork
Coomaghcheo, Curracahill, Adrivale
Cork
Boggeragh Mountains
Cork
Esk South
Cork
Carragraigue, Charlesfield, Inchamay North Cork
Glenlahan, Forehane, Cappaphaudeen
Cork
Glentanemacelligott, Glennakeel South
Cork
Taurbeg, Glasheenavargid
Cork
Taurbeg
Cork
Rockhill West
Cork
Coomagearlahy
Kerry
The Coom, Cordal
Kerry
The Coom, Cordal
Kerry
Knockauncurragh/Glanowen/Coom
Kerry
Tylagh
Kerry
Muingnaminane
Kerry
Tursillagh Expansion/Extension
Kerry
Cloghboola
Kerry
Pallas, Banemore, Cloghanenagleragh and Kerry
Kilfeighny
0.035
0.66
5.95
4.95
5.95
4
3
8.5
2
4.8
7
2
15.6
4.8
50
7
8
26
15
8
6.8
17.85
21
21.8
40.25
39
21
1
7
3
7
3
10
2
4
7
6
2
13
4
1
17
20
7
8
20
6
14
25
7
17
8
8
21
4
21
24
26
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
No
Yes
No
No
No
No
No
No
No
Yes
No
No
No
No
Kilpaddoge/Carhoonakineely
Kerry
Knockaveelish/Knocknalougha
Waterford
Beallough
Waterford
Nethertown & Shilmore
Wexford
Newtown/ Richfield/ Riesk/ Inish/ Ballyteige Wexford
Slob
23
15.3
1.6
-
28.5
12
2
1
43
No
No
No
No
No
Richfield
Grageelagh
20.25
3.6
7
-
No
No
Site ID
Site Location
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
County
Wexford
Wexford
131
Electricity Storage and Wind Energy Intermittency
43
44
45
46
47
48
49
50
Ballinoulart
Bolinrush
Knocknalour
Cappagh, Parkroe, Kilmore, Oldcastle
Moanvaun
Bunkimalta/Bauraglanna
Ballinveny/Borrisnafarney, Gortagarry
Skehanagh
Wexford
Wexford
Wexford
Tipperary
Tipperary
Tipperary
Tipperary
Tipperary
Site ID
Site Location
County
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
No
No
No
No
No
No
No
No
Under
Appeal?
36
14.45
16
0.66
42.5
7
5
Capacity
(MW)
24
17
3
11
1
17
6
5
Lacka
Tipperary
Ballybeagh
Kilkenny
Dromdeeveen
Limerick
Caherlevoy, Glengort South
Limerick
Tournafulla/Templeglentan
Limerick
Knocknasna
Limerick
Tooradoo
Limerick
Gortnagross
Limerick
Grouselodge
Limerick
Currachafoil
Limerick
Rosscurra, Aclare, Kilbranish North
Carlow
Moneypoint
Clare
Monmore South
Clare
Booltiagh
Clare
Cronelea Upper
Wicklow
Rin
Offaly
Pollduff
Galway
Derrybrien
Galway
Keelderry
Galway
Sonnagh Old, Kilchreest
Galway
Teevurcher
Meath
Teevurcher
Meath
Cuillalea
Mayo
Croughan West, Dooleeg More, Kilsallagh, Mayo
Knockmoyle, Laghtanvack
3
10.5
25.5
2.55
5.95
8
12
4.25
52.5
22.5
12.6
19.5
2.55
15
60.35
48
7
4.5
7.5
3.4
320
3
5
7
9
17
3
7
8
5
21
9
7
15
3
5
23
71
48
10
5
4
192
No
No
No
No
No
No
No
No
No
No
Yes
No
No
No
No
No
No
No
No
No
No
No
Yes
Yes
Alt, Bunnahowen
Mayo
Corrinshigo/Raragh (Clankee Barony)
Cavan
Gartnaneane
Cavan
Bindoo
Cavan
Ratrussan
Cavan
Edrans, Tullyco
Cavan
Mountain Lodge
Cavan
Artonagh, Tullyco
Cavan
Mountain Lodge
Cavan
Snugborough & Carrowmore
Cavan
Corrie Mountain, Tullymurray
Leitrim
Black Banks (ext)
Leitrim
Moneenatieve (Corrie Mountain ext)
Leitrim
Garvagh Glebe, Leckaun, Tullynamackduff, Leitrim
Bargowla, Seltan, Boleymaguire
3
3
15
77.5
3
7
48
26
37.5
10.5
3.96
6.8
5.1
32.5
3
10
31
2
32
26
7
8
6
13
No
No
No
No
No
No
Yes
Yes
No
No
No
No
No
No
Turbines
132
Electricity Storage and Wind Energy Intermittency
89
90
91
92
93
94
95
96
97
98
99
100
Cunghill, Lavagh
Carrownyclowan & Carrowmore
Lackan Townland
Cornacahan
Corkermore Hill
Meenacloghspar
Meenanilta
Meenahorna
Meenalaban
Meentycat
Cark Expansion/Extension
Cark, Newmills
Sligo
Sligo
Sligo
Donegal
Donegal
Donegal
Donegal
Donegal
Donegal
Donegal
Donegal
Donegal
Site ID
Site Location
County
101
102
103
104
105
106
107
108
109
110
Drumkeen
Ballystrang, Rareagh
Glackmore Hill, Three Trees
Sorne
Bauville Keeloges
Flughland
Glasmullan & Shandrim
Drumlough Hill
Baile Thoir
Arklow Bank
Donegal
Donegal
Donegal
Donegal
Donegal
Donegal
Donegal
Donegal
Donegal
Offshore
6.4
2.55
21
1.2
2.55
22.5
43.6
22.5
24.45
4.25
Capacity
(MW)
5
6
3
3
14
2
3
9
15
-
4.25
7.8
5
28
8.9
8
2
520*
9.1
5
16
3
5
8
13
2
200
Turbines
No
No
Yes
No
No
No
No
No
No
No
No
No
Under
Appeal?
No
No
No
No
No
No
No
No
No
-
133
Electricity Storage and Wind Energy Intermittency
134
Electricity Storage and Wind Energy Intermittency
APPENDIX 2. VARIABLE USED IN FINANCIAL MODEL
Variable
Cost of charging
equipment
Cost of storage equipment
Cost of regeneration
equipment
Operation and
maintenance costs
Input energy cost
Output energy price
Charging efficiency
Discharge efficiency
Hours of operation per day
Green energy credit
Cost of debt financing
Cost of equity or
mezzanine funding
Terminal value discount
Capital structure
Tax rate
Description
€ cost per kW for charging
equipment (for example
pumps or electrolyser)
€ cost per kWh for storage
medium (for example
water reservoir or tanks).
Modelled at zero loss
€ cost per kW for
regeneration equipment
(for example turbines,
engines or fuel cells)
€ cost (semi-scaled) for
plant operations and
maintenance
c / kWh for input energy,
typically the foregone
attainable price by the
wind energy producer
c / kWh for electricity
supplied to the grid
Efficiency of charging
equipment
Efficiency of discharge
Number of hours that the
storage system discharges
at rated power
Benefit accrued from
either ROC or similar
system
Cost of primary debt for
project
Cost of subordinated
capital
Appropriate discount
charge on future
production after 20 years
lifetime
Debt / mezzanine / equity
Tax rate for project
Typical Range
€ 400 - € 1,500 / kW
€ 1 - € 20 / kWh
€ 400 - € 2,000 / kW
0. – 0.5 c / kWh
0 – 4 c / kWh
3 – 7 c / kWh
40% - 90%
20% - 90%
5 – 20 hours
0 – 7 c / kWh
4–7%
7 – 12%
5 – 16%
80% primary debt
12% – 25%
135
Electricity Storage and Wind Energy Intermittency
136
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement