Energy Storage

4 - 7 October 2022

Infocus International
+65 6325 0210
enquiry@infocusinternational.com

Do not copy or redistribute without written consent of Infocus International and faculty
Infocus International is a global business intelligence provider of strategic information and professional services serving diverse business communities,
designed to provide insights and to assist our clients on the global stage.

We are a leading knowledge-management companies strategically based in Singapore, independently researching, producing and delivering market-driven
programmes globally (Asia-Pacific, Middle East, America, Europe and Africa).

Our company strongly believes that knowledge works as your global currency. We recognise our clients’ needs and respond with innovative, timely, and
results-oriented programmes. Our products are founded on high-value content in diverse subject areas, and the highest level of quality is ensured through our
intensive, accurate, and in-depth market research from local and international insights. Each programme is developed by a team of competent research
professionals who ensure our agenda provides detailed discussions and latest industry updates.

Infocus International is also widely recognised as a global leading conference and summit organiser. Participants in our events can look forward to high
value interactive sessions and enriching discussions among some of the most recognised personnel in their industries. A rewarding experience to delegates
is endeavored through innovative practices and excellent service.

Our training division provides extensive training programmes throughout the continents ranging from Finance, Energy, Telecoms & IT, Retail, Marketing,
Human Resource, Infrastructure, Mining, and Safety. These programmes are delivered by leading industry practitioners and field experts in a choice of
layouts such as workshops, and in-house programmes.

Our commitment to deliver high quality products and services is the main foundation of our reputation as a prominent player in the business intelligence
industry.

With our wealth of experience in organising conferences, seminars and training programmes, we welcome the opportunity to work with any corporation to
serve their particular industry.

For more information about Infocus International’s spectrum of activities, please contact Jessie Ang, General Manager
tel: +65 6325 0211 or fax: +65 6224 5090 or e-mail to jessie@infocusinternational.com
 Session 1:
Battery Storage
An example global growth forecast…

Forecast total of 358GW / 1,028GWh by 2030 (vs. 17GW / 34GWh, end 2020)
▪ Just over 50% will be to provide energy shifting of solar and wind generation
▪ About 25% will be at residential and commercial & industrial (C&I) scale
▪ LFP likely to have overtaken NMC for stationary storage for the first time in 2021
(and was ~30% cheaper)

Pumped hydro 2020

(159.5 GW)
 e.g. Germany: acceleration, led by Li-ion

Home systems:

Industrial systems:
 e.g. Germany: deceleration (business model?)

Large systems:

▪ From 2016 to 2019, systems were built almost exclusively for the provision of FCR
▪ During this time, FCR prices dropped significantly, due to the increasing saturation of
the market volume by battery storage
▪ After some technologically versatile installations from 2017 to 2019, lithium-ion LSS
have now been installed exclusively for the second year in a row
 Defining Storage “Capacity”

• Power plants are sized by power capacity

• Storage projects are sized by power and energy capacity
(where energy capacity determines duration)

Power : energy ratio

2:1 1:1 1:1

Energy
Capacity

Power
Capacity
Duration
Storage types & depths

Category Description Duration

Behind-the-meter, to
Distributed support or shape a Usually < 4 hours
customer’s own load

Aggregated (VPP)
Co-ordinated distributed Usually < 4 hours
installations, including V2G

Capacity, fast-ramp and

Shallow From 30 mins to 2-3 hours
ancillary services value

Daily load shaping and solar

Intra-day 4 – 12 hours
shifting

Covering renewable
Deep / long-duration ‘droughts’ or seasonal Days to weeks / months
smoothing
Battery Types include…

Lithium-ion (various): Flow batteries:

• Vanadium-redox
▪ NMC (NiMnCo)
• Zn-Fe
▪ LFP (LiFeP)
• Zn-Br
▪ LTO (LiTi)
▪ LCO (LiCoO)
▪ LMO (LiMnO)
▪ NCA (NiCoAl)

Other, e.g.:
• High-temperature sodium (NaS)
• Lead-acid
• Iron-air
• Zinc-air
• …

Images: Let’s Talk Science, International Flow Battery Forum, NGK Insulators
Battery prices: big falls, but levelling off?

(volume-weighted average
pack and cell price splits)
Li-ion battery cost breakdown

Where does the money go?

(data source: Bloomberg, 2021)
 Li-ion battery types

Cathode chemistry:

• LFP offer greater thermal stability and safety but lower energy density (energy per kg) relative to NMC
• NMC batteries have greater energy density (high nickel content) and higher power
• Early thermal runaway has been overcome by the addition of cobalt to the cathode
 Example comparison of Li-ion chemistries

Some key metrics

Other considerations include: safety, supply chain (materials), recycling etc.

Source: Stecca et al, Open Journal of the IES (2020)

A Battery System

Modules
Battery
Management
(A/C, fire etc.)

+
Ancillaries

Connectors Info: Voltage System

Monitoring Current
Communication Interfaces Temperature
Thermal management devices SoC
Cells SoH
Processed energy
DC …

Bidirectional
inverter

AC Energy Management System

System data monitoring
Cloud access
Transformer, Battery & inverter control
Switchgear etc.

‘Offtake’
 Containerised batteries: what’s inside?
Images:
Leclanché, Fluence

Containerised storage:
• Cells (few 10s of Wh)  Modules (few kWh)
 Racks (few 10s of kWh)
• 3 – 6 MWh in a typical 40’ shipping
container (li-ion)
 Total Energy Storage Costs
There are many…
(This was drawn for batteries, but equally applicable to any energy storage project)

Everyone tends to focus here...

How much revenue will it earn...?

Image source: Bushveld Energy

Ups and downs…

“Grid-scale energy storage systems are unlikely to see any price declines until 2024,
when manufacturing of lithium-ion batteries scales up to meet the increase in
demand from automakers”
Example storage costs… (US)
2021 U.S. utility-scale LIB storage costs for durations of 2–10 hours (60 MWDC) in $/kWh
Example storage costs… (US)
2021 U.S. utility-scale LIB storage costs for durations of 2–10 hours (60 MWDC) in $/kW
 BoS, Duration and Capex Metrics

Whereas power plants are capacity designed, batteries are both capacity and
duration designed.
- so a battery with the same power but longer duration costs more!

NB. Hypothetical costs example!

Degradation

“Cycle Life”: energy capacity drops to 80% of its initial spec.

“Lifetime” (e.g. warranty) depends on usage, e.g.:

▪ Depth of Discharge (DoD) per cycle
▪ State of Charge (SoC) range per cycle
▪ Cycles per day
▪ Rate of charge/discharge
▪ Average State of Charge (SoC)
▪ Temperature

“Calendar Ageing” occurs even if a battery is not cycled

Battery sizing considerations

Which strategy to account for degradation?

Oversize? Upgrade?

Reduced
Replace?
Battery capacity (& cost) Service?

Oversizing
Augment / replace
Contracted
capability

Reduced service

Time
Contract
length
 Degradation curves & battery warranties
▪ Storage manufacturers leverage lab test results to inform their warranty terms
▪ Differences in cycling, rest time, state of charge (SOC), temperature and other metrics all impact
the performance and degradation of battery cells
▪ There is no industry standard and degradation curves vary widely across lithium ion cell types
▪ Some warranties are becoming more restrictive to enable longer warranty terms (20 to 25 years)
▪ Some BESS providers offer more flexible warranty terms that put control and economic trade-offs
in the hands of the project owner or operator
▪ Flexible warranty terms typically include formulas instead of a fixed table of values that use
operational metrics as inputs (to derive degradation curves)

Source:
www.dnv.com/article/energy-storage-capacity-
warranties-beyond-the-fine-print-200339
 Typical battery warranty conditions

Term Metrics Notes

• Based on utilization: cycles or throughput
Capacity: (discharged MWh)
% of initial capacity in a table/schedule for
usage/cycling • Does PPA or biz model account for capacity loss?
each period of the warranty term • Augmentation or replacement strategy included
degradation
in service agreement or financial model?

• Even if BESS is underutilized, energy capacity still

Capacity: % of initial capacity in a table/schedule for degrades
calendar degradation each period within the warranty term • Annual or monthly regardless of utilization
• Include augmentation or replacement strategy?

• Non-linear energy degradation

Term of warranty e.g. 10, 15, or 20 years • Include augmentation or replacement strategy?

• Where and how often is T measured?

Temperatures Maintain ~25 °C across all cells • What constitutes ‘exceedance’ (spot reading or
average over time)?

Average state of charge • Ensure operations don’t exceed maximum

<50% annual average SOC
(SOC) allowed average SOC

e.g. 0.25 C, 0.5 C, or 1 C • Ensure operations don’t exceed maximum

C-rate
(i.e., 4 hr, 2 hr, or 1 hr battery discharge rate) allowed c-rate

Second-by-second BMS data, later aggregated • Real-time access

Data storage
into minute/hourly records • Secure, offsite storage

• Liquidated damages ($ loss) if exceeded?

Response & repair times Day(s) for response, weeks for repair • Onsite spare parts?
Insurance, guarantees & financing
Example: the Cellcube “bankable battery guarantee”
▪ Enerox (CellCube) have a contract with Munich Re
▪ Enerox can offer their vanadium redox flow batteries with a bankable
insurance to guarantee product and performance accuracy
▪ Customers can choose this long-term performance warranty from
Munich Re for 10 years
▪ A re-certification can prolong coverage for another decade

▪ What’s covered?
▪ continuous operation level
▪ insurance against malfunction
▪ repair or replacement costs
▪ problems in workmanship
▪ manufacturer’s potential insolvency

Enables project financed projects such as the Bushveld Vametco mining

solar and storage minigrid, Brits, South Africa (1MW/4MWh)
Flow batteries
Claimed advantages:
• 100% DoD without degrada on (→ longer lifespan)
• Can be discharged to zero volts without damage
• Can stay fully or partially discharged for prolonged periods
• More abundant source materials
• No risk of thermal runaway or fire

Key disadvantages:
• Lower efficiency
• Small-scale (and low growth)
• Technology & supplier risk
 Flow battery: still small scale?

▪ Redflow: Zn-Br flow battery maker

▪ 0.5 MW, 2 MWh installation in San Bernardino County, California
▪ Redflow’s “biggest installation to date”
▪ Est. cost: US$1.2 million
▪ 12 x 160 kWh ‘Energy Pods’
▪ 4 strings tied to 4 x 125 kW inverters
▪ Site, a bioenergy facility, already has 2 MW biogas-fuelled microgrid
▪ Battery will reduce peak energy usage (4 - 9pm)

▪ Son Orlandis PV farm: 3.34 MWp (Phase 1)

▪ 5 MW, 20 MWh VRFB project
▪ Phase 2: 1.1 MW / 5.5 MWh VRFB
▪ “US’ biggest VRFB project so far”
(All Dec 2021) ▪ Expected to be operational by the end of 2023
Flow battery, tidal power & hydrogen

▪ At the European Marine Energy Centre (EMEC) hydrogen R&D facility

▪ 48 Invinity VS3 battery modules (i.e. 37.5 kWh x 48)
▪ The battery will ‘smooth’ the power from the tidal generation to ensure EMEC’s 670 kW hydrogen
electrolyser has a stable, renewable power supply
▪ Full demonstration of the integrated system due in autumn 2022
 Best of both? (Hybrid battery systems)
Example: Energy Superhub Oxford (ESO), UK
▪ Vanadium flow battery from Invinity Energy Systems (5 MWh)
▪ Wärtsilä lithium-ion battery (50 MWh)
▪ System power capacity 48 MW
▪ Developer: Pivot Power
▪ Fully energised in December 2021
▪ Flow battery acts as the first line of response
▪ Li-ion battery only called after the required response exceeds the capacity of
the flow battery
▪ Aim is to reduce the degradation of
the Li-battery
▪ The site will have EV chargers:
▪ 10 x 300kW, 16 x 7-22kW and
12 x 250kW Tesla Superchargers
 Many other emerging batteries…

▪ First pilot due online in late 2023

▪ Little detail: an "aqueous air" battery system
▪ “Safe, cheap, abundant” materials
▪ 1 MW, 150 hours ‘replicable block’
(Of course any battery can last 150 hours if you discharge it slowly enough!)

• ESS Inc. Iron flow batteries (“iron, salt and water”)

• “Geared toward 24-hour storage or longer” • SB Energy is part of SoftBank
• Initial market entry targeting applications • Deal is for 2GWh up to 2026
offsetting diesel generator usage • First one will be at a solar power plant in California
• “Storing electrical energy within zinc metal” • Energy storage “up to 12 hours”
(precipitation & dissolution in a KOH electrolyte)
• “More than 20,000 cycles” and “life expectancy of 25 years”
• 10-yr extended warranty (backed by Munich Re)
• Another deal with Enel, Spain: 17 shipping container-sized
batteries, adding up to 8.5 MWh
Safety is an important issue

▪ For permitting, stakeholders, liability & other factors…

Buyers are forcing ethical battery purchasing

▪ Clean Power Alliance, a “community choice aggregator” is one of the biggest buyers of
renewable energy and battery storage output in the United States
▪ It has banned forced labour, child labour, slavery and other human rights abuses in 3
long-term PPAs
▪ Two contracts are 15-year, one is 20-year
▪ The contracts add up to 206.5 MW of solar PV and 163 MW/652 MWh of battery
storage
▪ Due online from late 2023
▪ An expanded supply chain code in the contracts includes abuses in the mining,
processing and procurement of materials used in the production of both batteries
and solar panels

93.5 MW solar + and 71 MW/284 MWh: Arica facility in Riverside County, Calif. (15 yr)

65 MW solar + 52 MW / 208 MWh: from the Daggett Solar Project in San Bernardino County, Calif (15 yr)

48 MW solar + 40 MW/160 MWh: Resurgence Solar II project, in San Bernardino County, Calif. (20 yr)
EU battery regulations (2020)

New batteries made in the EU from

2030 will require a minimum recycled
content of 12% for cobalt and
4% for both lithium and nickel

Legislated metal recovery rates by

2030 are 95% for cobalt and nickel
and 70% for lithium
Recycling in China

“Existing technology in China can theoretically

recover around 80% of the components of
different battery types
but
Currently, estimates sourced in Chinese media
report that only around 30 – 40% of battery
materials are being recycled”

• In 2021, Beijing issued the country’s 14th Five-Year Plan (2021-25)

• This laid out the goal of building a more complete battery recycling system by 2025
Recycling demand – not enough? (for now)

“It could take 15 years or more for old batteries to eventually arrive at recycling plants,
and in some cases as long as 25 years”

“While the industry waits for early models of EVs to hit junk yards in big numbers… by
2025 there may be three times more recycling factory space than scrap to run the plants”

“In 2025, 78% of the available scrap supply will be coming from manufacturing waste,
while end-of-life batteries will account for 22%, according to new research by Benchmark
Mineral Intelligence”
An Example ‘RFP’ for Battery Storage

e.g.
 Session 2:
‘Utility’-scale energy storage
Balancing the grid
 System inflexibility markers

e.g. the UK in 2021:

▪ Rising gas prices and tighter margins
▪ Day-ahead energy prices reached
over £1,800/MWh
▪ Balancing prices reached
over £4,000/MWh
▪ The 10 most expensive days ever in the balancing mechanism
▪ Total balancing costs £860m higher than in 2020
▪ Wind power was curtailed on 54% of days (>2.3TWh of energy in total)
▪ Curtailment was mainly due to network constraints

▪ Prices for Dynamic Containment (DC), a new set of frequency response

products, remained high (at the price cap of £17/MW/h)
▪ 3.3GW of new battery storage in the Capacity Market (CM) T-4 auction for
delivery in 2025-26, at a record price of £30.59/kW-yr (derated)
Frequency and ‘reserves’
Storage vs. Conventional “Reserves”
100 MW storage vs. 100 MW gas turbine

Charging = Demand ‘Turn-up’

 The Hornsdale Power Reserve, Australia

aka the Tesla “Big Battery” (but owned by French developer Neoen)

▪ From tender to operation within a year.

▪ Sited alongside an existing 309MW wind farm
▪ Original capacities: 100MW, 129MWh
▪ 70MW (for approx. 10 min duration) contracted to SA
government for grid stability (“FCAS”) services
▪ Remainder available for other merchant revenue streams:
e.g. FCAS markets, wind farm load shifting, arbitrage

▪ Can charge from the grid when no wind

▪ Built with equity finance
 HPR: a profitable project!

▪ Estimated capital cost: $A90m

▪ 70 MW network services contract with South Australia government:
▪ $A4m per year for 10 years (EBITA)

▪ Started operating in December 2017:

▪ In H1 2018*:
▪ Earned $A2m from the government contract
▪ Earned $A10.8m from market revenues
▪ Most market revenue ($A7.1m) came from the FCAS market.
▪ Helped Neoen reduce its own FCAS costs from wind farm operations

▪ i.e. potential to earn >A$25 million in revenues in its first calendar year.

▪ In Q1 2020:
▪ Neoen’s total storage revenue was A$ 36.3m (almost all from HPR)

*Source: Neoen public offering brochure (2019)

HPR: Expansion

50MW/64.5MWh expansion completed

in September 2020
▪ $73 million expansion:
▪ $8m ARENA grant, $15m state funding,
$50m project financing (CEFC)
▪ The full 150MW now includes Tesla’s
“Virtual Machine Mode”
▪ This enables the battery inverters to
emulate inertia services to the grid
▪ HPR could provide up to half of South
Australia’s inertia needs
Inverter types

Grid-following Grid-forming

Synchronises to the grid waveform, Sets an internal waveform reference,

adjusting power to ‘follow’ the grid adjusting power to maintain this

Requires an external reference signal, Can keep operating at predictable

and shuts down if this is lost power without external support

Limited response speed and Instantaneous power output adjustment

can exacerbate disturbances can stabilise the grid (synthetic inertia)
Largest grid-forming battery

Nov 2021:
▪ $180 million Torrens Island battery, South Australia
▪ Owner AGL, battery provider Wärtsilä
▪ Potential expansion to four-hour duration in the future (1000 MWh)
▪ Operational early 2023, initially in grid-following mode
▪ Will use >100 inverters supplied by SMA
▪ Will become grid-forming “once Australia has finalized the regulatory landscape”

Grid-forming technology was flagged by

the Australian Energy Market Operator
as “a top priority”
 e.g. UK’s ‘Dynamic Containment’ service

About the DC service:

▪ Requires participants to provide power in <1s when frequency deviates by more than 0.2Hz
▪ There are both a low frequency (DC LF) and high frequency (DC HF) service
▪ Operates as a pay-as-clear auction for 4-hour (EFA block) contracts
▪ A previous price cap of £17/MW/h has been revised to vary by EFA block
▪ The requirement is set by the largest single generation loss on the system and level of
forecast inertia
▪ NGESO’s median requirement is around 600MW, with a maximum requirement of 1600MW
▪ Although lucrative in 2021, the service is likely to be saturated this year

Also:
Pre-fault Dynamic Regulation (DR) and Dynamic Moderation (DM) services launched in 2022:
▪ DM respond within one second, sustain response for at least 30 mins
▪ DR respond within ten seconds, sustain response for at least 60 mins
 Solar + Battery for Grid services (and more)
Clayhill
▪ UK’s first “subsidy-free solar farm” (built 2017)
▪ 10 MW solar PV, 6 MW / 6 MWh battery (5 x 1.2 MW)

▪ Built by Anesco
▪ Added to a ‘VPP’ operated by Limejump (2018)
▪ used for dynamic frequency response & balancing market revenues

▪ Industry-first “floor price” agreement with EDF Energy and Upside Energy (2019)
▪ Using their trading platform for Capacity and Firm Frequency Response (FFR) markets, plus
wholesale market arbitrage

▪ Sold to Gridserve (2020), with Anesco signing a 20-yr service agreement

▪ Gridserve will use it to provide zero-carbon electricity to a nearby EV charging hub

The solar farm wouldn’t have been built without the battery,
which adds grid services revenue to the solar energy sales
 Batteries & Frequency Control Prices
Large-scale storage systems by application:

(Example: Germany)

FCR auction prices: (Frequency Containment Reserve)

(Figgener et al 2020, Journal of Energy Storage)

Price signals – volatility!
(Extreme) expensive Peaks…

8th, 9th & 10th: max. daily temp > 40°C

Data source: aemo.com.au

HPR and arbitrage: $1 million in 2 days?

http://reneweconomy.com.au/tesla-big-battery-moves-from-show-boating-to-money-making-93955/
Peak Shaving & the 3-4 hour “rule”

1,790 MWh 1,002 MWh 583 MWh

(450MW x 4hr) (350MW x 3hr) (200MW x 3hr)
Natural gas peakers – big capacity, little usage

Capacity factors of natural gas OCGTs in the US, 2019:

~60% of open cycle gas

turbines (OCGTs) in the U.S.
never ran for more than six
consecutive hours in 2019

~80% of the U.S. gas peakers

had capacity factors below
15% in 2019
 Battery systems replacing natural gas

e.g. South Bay/Moss Landing (California)

▪ PG&E approval for four battery plants to provide local capacity
▪ Replaces three “reliability must run” gas generators (Calpine)
▪ “Estimated benefit of $233m over 10 years” (source: CPUC)

▪ Includes two of the largest battery systems ever:

▪ 300 MW / 1,200 MWh (Vistra Energy)
▪ 182.5 MW / 730 MWh (Tesla).
▪ 20-year “resource adequacy” (capacity) contract with PG&E
(plus less secure merchant revenues)

▪ Big scale:
▪ lowers component costs
▪ allows spare capacity for other revenues: CAISO markets (energy, grid services)
▪ Sites: Tesla system on PG&E land, Vistra in existing turbine building
▪ Vistra system uses existing interconnection (from old gas plants)
Solar + storage: a growing trend

Source: https://www.eia.gov/todayinenergy/detail.php?id=49756
Maximising output: avoiding ‘clipping’

If batteries are behind the inverter (DC-coupled):

▪ Analysis for one US utility suggested avoided peak clipping could be almost 10% of a
battery’s revenues*

*Very dependent on sizing arrangements (PV, battery, inverter)

Source: The Brattle Group (2019)
 Example Dispatchable PV (Hawaii)
Kauaʻi Island Utility Cooperative
▪ State’s first dispatchable PV plant (March 2017)
▪ 17 MWp PV and 13 MWAC, 52 MWh battery storage.
▪ Batteries are Tesla Powerpacks.
▪ Will displace 1.6 million gallons of diesel.
▪ Designed to avoid curtailing solar at midday and feed
into the grid during peak hours (in the evening).
▪ 20-year contract at 13.9 $c/kWh
(less than the cost of oil at the time).

▪ Hawaii's renewable goal is

100% by 2045.
 Competitive peak-time solar in mainland US

e.g. Arizona Public Service (buyer) & First Solar (seller)

▪ 50MW/135MWh (2.7 hr) battery system and 65MW PV.

▪ Won an RFP by APS for peaking capacity resources

▪ Addresses “duck curve” & electricity during peak demand hours: early evening (3-8pm).
▪ "We don't want midday energy. Sometimes it creates bigger operational problems than it solves."
▪ Due online in 2021 with a 15-year PPA

▪ PV plant charge the battery during the day and deliver power until sunset, allowing the
whole project to qualify for the ITC (30%)
▪ During peak, APS has full use of the 50-MW battery (maximize capacity until discharged).

▪ APS is not buying outside peak hours:

▪ First Solar could explore other revenue streams in the off-hours

(it must ensure capacity to deliver its obligation when the peak starts).
 Utilities making the case for Storage
When Utilities don’t want energy!

i.e. PPA will specify this!

(source: APS 2017 “peaking capacity” RFP)
NV Energy Solar + Storage PPAs (Dec 2019)

Designed specifically to attract solar + storage:

▪ Price during system peak hours (4 to 9 pm) 6.5x the price at other times

Awarded PPAs:

Project Online date Solar capacity Storage capacities

Gemini Dec 2023 690 MW 380 MW, 1416 MWh

Southern Bighorn Sept 2023 300 MW 135 MW, 540 MWh

Moapa Dec 2022 200 MW 75 MW, 375 MWh

TOTALS: 1190 MW 590 MW, 2331 MWh

Source: The Brattle Group (2019)

Spanish auctions: “management capacity”

The price received by power plants depends on:

▪ the auction awarded price
▪ the market price
▪ the “market adjustment percentage”

Price = Price Awarded + Adjustment(%) * (Market Price – Price Awarded)

For “Facilities with management capacity”, the adjustment is 25%

For “Facilities with no management capacity”, the adjustment is 5%

What is “management capacity”?

▪ Installations which have a storage system able to store “an amount of energy equal to or greater
than the result of multiplying the power of the installation by two hours”
▪ They are only allowed to store energy produced by the plant, not energy from the grid
▪ Such projects have a greater exposure to market prices
▪ The objective is to stimulate energy production in the most expensive hours of the day
Not enough grid?

(MISO territory, US, 2016-2020)

Curtailment
e.g. California:

Source: Energy Information Administration, US

 Virtual Power Lines / Non-Wire Alternatives
130 MW generation & peak demand,
through a 100 MW power line?

Over-supply relief:
e.g. Terna, Italy
▪ ~500 GWh of curtailed wind energy in
2010 in the Campania region.
▪ Terna installed three NaS batteries with
total capacity 34.8 MW/250 MWh.
▪ Investment: €160m (in 2014)
▪ Stored energy transported whenever
transmission lines not congested.
▪ Batteries also used to provide primary
and secondary frequency regulation.

Source: IRENA (2020)

Mobile batteries: Power Edison
▪ Power Edison have “a GW-scale pipeline of projects with utilities”
(identities undisclosed)
▪ Biggest project so far:
▪ 200 MW heavy-duty EV charging hub at a New Jersey seaport
▪ Mobile batteries enables a site to be more flexible in matching storage to
shifting charging demands
▪ If and when permanent grid infrastructure is expanded to meet a site’s
charging needs, the batteries can be moved to another location

EV charging is one growing

example of where flexible
battery deployment can offset
T&D upgrade timing uncertainty

Changing seasonal needs

(i.e. stress locations) are
another potential use case
 Innovative or Insane?

(Sept 2021)

▪ Grid-scale battery company PowerX

▪ Plan to build “autonomous battery ships to transfer
offshore wind electricity” (Power Transfer Vessels”)
▪ First ship “in 2025”
▪ 100 grid-scale batteries (~200MWh) per ship
▪ Japan is surrounded by deep water & prone to earthquakes
▪ Undersea cables will be even more expensive than elsewhere

▪ “Electricity can be generated where it is most cost efficient

to build an offshore wind farm and transferred anywhere in
the country”
▪ Range expected to be around 300km

▪ Gigafactory to be built in Japan:

▪ annual production capacity of 1GWh by 2024, scaling to 5GWh by 2028
 Session 3:
Distributed & behind-the-
meter energy storage
‘Grid edge’ technology & disruption

Consumers  Prosumers

Efficiency
DSR
Decentralisation Generation (PV)
Storage
Microgrids

Smart meters
Smart sensors
EVs
Smart appliances
Chargers & V2G
Heat pumps
Electrification Digitisation IoT
SCADA
Smart heaters
Aggregation (AI, ML)

Carbon reduction & Open, real-time communications,

distributed generation automated operations
 What can BTM batteries do?
BTM = “Behind the Meter”

Source: IRENA (‘Behind the Meter Batteries’, 2019)

Distributed vs. centralised storage?

February 2019, ISO-NE (US):

▪ Sunrun (residential solar and storage company) won a forward capacity auction
▪ 20MW bid of aggregated residential solar + storage systems
▪ This represents ~5,000 systems
▪ Revenue will be shared among households instead of to one power plant
▪ The systems also directly lower bills for households and provide backup power
to families in case of grid outages

• In California, SCE has also used a portfolio of storage instead of a 262 MW gas peaker
• The portfolio include a big battery (100MW) but also 14MW of aggregated domestic
systems (from Swell Energy).
Evolution of residential BTM models

Early adoption: hardware sale

Drivers: individuals seeking to use more of their own PV and reduce grid
imports (with poor export & high import prices)


Managed service
Driver: convenience of having utilities manage customers’ import/export and
bundling into a flat-rate “cloud” service


Accessing new revenue streams
Driver: lowering flat-rate pricing through new revenue streams (grid services)
by managing multiple customers’ import/exports (aggregation)
 The role of Smart Tariffs
Example:

• First tariff to offer half-hourly prices, reflecting actual wholesale costs

• When wholesale prices go negative, consumers are paid to consume
• Half-hourly energy prices set daily at 4pm for the following 24 hours
 ToU tariff impacts?
e.g. EV charging and dynamic pricing (Octopus Agile):
▪ Providing market benefits without active external involvement
▪ EV drivers are highly responsive, e.g. Octopus ‘Agile Energy’:
▪ reducing peak time energy consumption by 47%
▪ Saving £132 per year compared to Octopus’ standard tariff

More examples of time-based tariffing:

▪ Austin Energy created a “free energy” off-peak tariff for EV charging, which successfully
moved virtually all EV charging outside times which would cause grid congestion impacts
▪ South Australia Power Networks’ “Solar Sponge” tariff offers a 25% lower tariff from
10am to 3pm, to absorb excess solar
 Aggregation: opening up new flexibility supplies

Small, distributed resources Large-scale service requirements

Image: Kubli & Canzi (Renewable and Sustainable Energy Reviews, 2021)
New England: ‘ConnectedSolutions’

e.g. New England ‘Connected Solutions’

▪ Incentivising managed battery storage

Key program aims:

▪ Reduce peak power costs for consumers:

▪ 10% of the highest demand accounts for 40% of customer spending
▪ De-risk investment (make storage ‘bankable’):
▪ Reliable, contractual revenue streams
▪ Definition of standardized eligible systems
▪ Support bigger and longer-duration batteries

▪ Align customer battery discharges with regional demand peaks:

▪ Focus on wider grid system benefits
▪ Give utilities a way to manage BTM storage without having to own it
New England: ‘ConnectedSolutions’
The Contract:
▪ Between the customer (battery owner) and their utility (or a third-party aggregator)
▪ Multi-year, pay-for-performance
▪ Summer incentive rate is ~$225 per kW per event
▪ Incentive rate is guaranteed for 5 years
▪ All customer types (C&I, residential) are eligible

Business model features:

▪ Attractive returns for consumers
▪ All ratepayers gain, due to increased grid efficiency
▪ Regional peaks are longer (~3hrs) than customer peaks (15 mins),
so longer-duration batteries are favoured
 The Tesla VPP, South Australia

Eventual aim is to create a VPP of up to 250MW of battery capacity

▪ Each customer site consists of three key system components:
▪ 5kW Solar PV system with associated inverter
▪ 5kW / 13.5kWh Tesla Powerwall 2 – AC coupled
▪ Powerwall 2 Gateway to manage site behaviour
and communications

Operating objectives:
▪ In ‘solar shifting’ mode, the battery consumes surplus PV (maximise self-consumption)
▪ The battery tries to avoid grid import outside solar times (by discharging to the home)
▪ Where necessary, the battery will pre-charge from the grid:
▪ to optimise for Time-of-Use tariffs
▪ to manage state-of-charge for VPP dispatch

▪ For VPP operation:

▪ Usage includes: wholesale market arbitrage, FCAS and integration with SA Power Networks’ ‘Solar Sponge’ tariff
▪ Total export (battery + PV) is limited by a static grid limit of 5kW
▪ A pilot programme for 5MW of VPP showed that export capacity could be increased to 6-8MW through ‘dynamic operating
envelopes’
Source: Advanced VPP Grid Integration (SA Power Networks), May 2021
Tesla VPP in California (with PG&E)

June 2022:
▪ “The largest distributed battery in the world”:
▪ There are ~50,000 Powerwalls that could be eligible, adding up to 500 MWh of energy capacity
▪ Opt-in to the Tesla Virtual Power Plant (VPP) with PG&E and your Powerwall will be dispatched
when the grid needs emergency support
▪ Through the Emergency Load Reduction Program (ELRP) pilot, participants receive $2 per kWh
delivered during an event
▪ Depending on the events and the number of Powerwalls homeowners have, they could earn
“anywhere from $10 to $60 per event”
▪ Customers can adjust their contribution, while maintaining backup energy for outages
VPPs for ancillary services (US)

▪ Will participate in the PJM Interconnection wholesale market

▪ 0.55 MW / 2.2 MWh VPP scheduled to be operational Q1 2022
▪ Homes can back up power supply in case of outages
▪ Local network can free up hosting capacity for distributed energy
resources (e.g. PV, EVs)
▪ Sunverge’s control platform can simultaneously run multiple services
with different priorities

▪ Jan 2021: VPP contract approved by the Hawaii Public Utilities Commission
▪ $25 million contract for capacity and fast frequency response grid services
across Oahu, Maui and Hawaii Island
▪ Swell will aggregate ~6,000 mostly residential solar-powered batteries
Germany: EVs dominate battery capacity

Stationary + mobile
battery storage 

 EV growth
EV Charging terminology

V0G Behavioural control (usually ToU tariffs)

One-way

V1G Externally managed charging, via communications link

V2H EV as a source of power to the home (e.g. backup1)

V2B EV as a source of power to a building

Bi-directional

V2X EV as a source of power to other electrical appliances (e.g. off-grid, camping etc.2)

V2G EV able to export into the grid

VGI EV more closely integrated into the grid and its operations

1 Onestudy has shown that a 32 kWh EV can provide 300 hours of continuous backup
household supply when combined with solar generation
2 In effect, EVs can be used to transport energy to locations which lack electricity access
Vehicle-to-…

Source: electrical-installation.org
 Ford F-150 (V2H as supply resilience)

▪ “Intelligent Backup Power” can feed power through a hardwired wall

charger into the home's electrical system
▪ Up to 9.6 kW of power output
▪ Energy capacity ~110-150 kWh
▪ vs. daily average consumption in the US of about 30 kWh

▪ Can automatically switch on when there's an electrical interruption

▪ Feeds power from the battery pack to an 80-amp, 240-volt home charger,
then to an inverter to convert to AC
▪ Will also be able to power your house during the day (when electricity rates
are higher), then recharge overnight
‘Full’ V2G examples

e.g. UK:
▪ The world’s largest residential V2G trial
▪ Over three years, with more than 320 chargers
▪ Trial found that V2G could earn consumers GBP 340 per yr through tariff
optimisation, compared with GBP 120 when using one-way smart charging
▪ Also provided firm frequency response (FFR) and dynamic containment(DC):
▪ Annual revenues rose to GBP 513 with FFR and GBP 725 with DC

e.g. Germany
▪ A pilot project showed EVs could address transmission grid congestion
▪ Wind power in northern Germany was used by EVs in that region
▪ Electricity from charged EV batteries was fed back into the southern and
western grids, replacing fossil fuel-based power
 C&I: no “typical” BTM storage case…

Example: McKinsey study into C&I user deployment:

“Identical buildings next door to each other can have

entirely different patterns of electricity use”
(So averaged data isn’t helpful in identifying profitable customers)

Energy-storage-dispatch models need to consider real-world data on:

▪ Electricity production and consumption (“load profiles”)

▪ at intervals of seconds or minutes for at least a year

▪ Battery characteristics, including price and performance

▪ Electricity prices and tariffs

http://www.mckinsey.com/business-functions/sustainability-and-resource-productivity/our-insights/the-new-economics-of-energy-storage (2016)
 …But the Key Elements are Common
Maximise usage & value
Energy Usage of onsite generation
• Fixed, $/kWh
• Variable, $/kWh (varying with times and/or tiers)
Shift energy to avoid
+ peak time imports

“Bandwidth” Requirement
• Demand charge, $/kW (based on maximum kW) Shift energy to
minimise peak demand
+

Energy Delivery (Grid Costs) Minimise cost of grid

• Fixed charges usage (volume, timing)
• Variable charges (pass-through grid costs)
Make onsite low-carbon
= generation dispatchable

Total Electricity Bill + Fuel Bill (for onsite generators)

+ Cost of Outages (remote sites)
 Reducing energy bills (power & energy)

• Reduced grid capacity requirement

• Reduced demand charges

Source: https://solartechonline.com/commercial-energy-storage/

Reduced peak-
rate energy costs
 Demand Charge Reduction
Utilities’ revenue from C&I electricity sales in the US is composed of:
• ~3% from fixed customer charges
• 25% from demand charges
• 72% from energy charges.
The split varies moderately by region but significantly by customer.
(Demand charges >50% of a monthly bill are common).

The impact of PV depends on timing! (Full-year Load & Weather data is needed)

Source: NREL, 2017

 EaaS / ESPCs (Energy Savings Performance Contracts)
Key Features
▪ A mechanism for monetising efficiency
▪ Removes up-front cost for the energy consumer
▪ Contract is for an operating lease on new technology
▪ The efficiency savings are big enough to pay for the lease plus an excess
▪ The technology provider’s project is financed from a 3rd party, based on the lease contract length
▪ At the end of the lease, the technology typically remains with the customer
(who get the full savings but need to self-O&M)
▪ For battery storage, a contract structure may well be different,
e.g. asset management or service agreement, retained battery ownership etc.

Example contract elements:

▪ Guarantee on the value of energy savings
▪ Monitoring to verify savings
▪ Monthly fee &/or share of cost savings
▪ Duration (typically 5 to 10, but up to 20?)
▪ Usage for grid services stacking
 Example: C&I Storage (as a Service)

“This 100% renewable energy solution enables

corporates to save up to 20% on their energy costs Example client requirements:
by installing redT flow machines alongside solar PV. • Site ownership for at least 25 years
The solution also enables customers to hedge • Investment grade credit rating
against rising energy prices as part of a 25 year solar • Available land/roof space
+ storage corporate PPA provided by Statkraft…
• Average demand load >150kW
… and access an innovative VPP to maximise value • Must provide half-hourly demand
streams from flexibility and optimisation services. profile, details of tariff & export
connection.
There are no upfront capex or ongoing
maintenance costs”.

Q: What are the pros & cons (from supplier & client sides?)
 Commercial V2G example

• An initial trial will use the batteries of 28 buses,

capable of providing 1MW of capacity
• From 6pm to 7am the buses are available for energy
trading and flexibility services, consisting of either
demand side response (DSR) or V2G services

• Bus2Grid project is funded by the UK government

• “If the entire London bus fleet of around 9,000 vehicles was
converted to V2G, it could provide 150,000 homes with
electricity”

• At 300kWh per bus and 40 kW export capacity per charging

point, a 120 bus depot (London average) could provide:
• 36MWh and 4.8MW per depot
Distribution-grids: storage & load-smoothing

Transformers • Designed to operate most efficiently ~70% of rated power

& • Battery storage can shift loads to keep transformers operating
System Losses more efficiently

‘Overload’ events • Battery discharges if demand at the transformer > upper setting
&
Peak Shaving • Batter charges if demand < base setting

990 kVA rating

power

energy

Image source: NREL (2020) for USAID/Gov of India

 What size battery?
Plotting overload events
for a particular substation:

Most
extreme i.e. 4-hour
event battery
Energy (kWh)

e.g. battery sized (Also factor in

for 70% of events any expected
load growth)

Power (kW)
Most events fully mitigated,
more unusual events partially mitigated
Image source: NREL (2020) for USAID/Gov of India
 Solar & BTM Storage
Using BTM aggregated storage to address reverse power flows & peak-shaving
Example DNO-led trial (Northern Powergrid, UK):
▪ 40 domestic scale Moixa battery systems in 36 households
▪ 27 also had PV installed
▪ The trial ran from 2017-2019

e.g. Maximum Impact scheme:

▪ Batteries empty before generation,
full before evening peak
▪ Batteries forced to (dis)charge this way:
Reduce excess demand (65%) and
& peak export (38%)

Seasonal considerations:
▪ Summer: low evening demand meant SoC did not drop significantly (not enough capacity to store
excess solar energy the next day)
▪ Winter: low levels of PV (not enough battery charge to shave the evening peak)
 Solar & storage on local networks

• Network topology?
• How much PV?
• How much local load?
• Grid capacity?
• …

Example of modelled scenarios

With & without PV: With PV, with & without storage:

overvoltage
 Community Energy Storage

▪ New South Wales, Australia

▪ A two-year trial on parts of the grid with lots of solar
▪ Will soak up consumer generated solar in the day
▪ Will discharge at peak (evening) times,
smoothing demand (and cutting costs)
▪ Also used to help stabilise the local grid
▪ The pictured battery is 150kW, 267kWh
▪ Two other locations are planned
▪ Each battery costs ~$400,000*
▪ Forecast annual savings of $100 - $300 per year to consumers

*1 AUD = 0.78 USD (Feb 2021)

 Community Energy Storage

▪ United Energy (distribution company in Victoria, Australia)

▪ $11 million ‘Electric Avenue’ project
▪ 40 batteries built and mounted to electricity poles across
United Energy’s low-voltage distribution network
▪ They will operate as a virtual power plant (VPP), from 2023
▪ Each battery: 30kW / 60kWh
▪ “Capable of powering up to 75 homes for at least 2 hours”
▪ Aim: “Soak up rooftop solar generation, while also
reducing peak demand and addressing grid stability issues”
▪ Mounting on poles avoids land constraints
▪ They “mimic the appearance of transformers”
 Session 4:
Long-duration and alternative
storage solutions
Low-carbon system flexibility options
How much storage do we need (& why)?

Grid services & integration

Fast response <1 hr
(frequency, voltage, balancing)

Peak demand
Intra-day <10 hrs
(capacity planning & cost)

Storms & weather patterns

Multi-day <200 hrs?
(capacity planning)

Overcapacity
Seasonal <1,000 hrs?
(capacity planning)
 Renewables & ‘Variability’ (hours & days)
“Dunkelflaute” (‘Dunkelheit’ (darkness) & ‘Windflaute’ (little to no wind)

(Germany, January 2017)

Filling a wind lull?
 Variability (seasons)

Demand:

Supply (solar):

(UK data, Staffel & Pfenninger, Energy (2018)

Meeting UK winter peak demand

5pm on a dark, still February evening…

 Capacity growth  surpluses?
An example summer week in the UK: Data from www.gridwatch.templar.co.uk, 2017

GW

Wind x 5, Solar x 2:
• Excess Energy = 1,130 GWh
• Max Excess Power = 20.9 GW
Rising demand for ‘dispatchable resource’

Australian Energy Market Operator’s (AEMO’s) 2020 Integrated

System Plan anticipates scenarios requiring, by 2040:
26 to 50 GW of new large-scale renewable energy generation
supported by
6 to 19 GW of new dispatchable resources
(pumped hydro, large-scale and distributed batteries, VPPs and other
‘demand-side’ participation)

“The most pressing utility-scale need in the next decade,

according to AEMO,
will be for 4-12 hour storage
to manage variation in output from renewable generation
and meet demand as coal leaves the market”

Source: Clean Energy Council (2021)

Storage on what scale?

One UK study concluded:

90% of energy storage will be in the 4 – 200 hours range

There would be little need for long duration storage with a

mix of ~80% wind and 20% solar
(which roughly matches seasonality of demand)

Having a some overcapacity in renewable generation

(10-15%) is very cost-effective, because it
reduces energy storage requirements

As generation costs reduce, higher proportions

of over-generation become appropriate
‘Long-duration’ energy storage (LDES)

Desired features of long-duration energy storage:

▪ Low marginal costs of storing additional energy
▪ Decoupling energy storage capacity from charge/discharge rates
e.g.
▪ charge with a compressor
▪ discharge through a turbine
(two different processes with different efficiencies)

▪ Widely deployable
▪ e.g. retrofits (gas storage to CAES, coal plant to TES)

▪ Scalable, e.g.:
▪ Few geographical requirements
▪ Modular
▪ Not dependent on rare materials

▪ Low lead-times (vs. T&D grid upgrade and expansion)

Source: LDES Council (2021)
Large-scale energy storage options

Current Emerging / Future

Fully commercial, large scale: Demo / pilot:

 Batteries (Li-ion, flow)  Mechanical/gravity (various)
(e.g. Moss Landing, US: 300 MW & 1,200 MWh)  Thermal (various)
 Pumped hydro  Batteries (various)
(e.g. Coire Glas, Scotland: 1,500 MW & 30,000 MWh)

Limited / pre-commercial: Future?

 Batteries (new chemistries)
 Compressed air
 Power-to-H2 (+ Salt cavern, depleted gas field)
 Liquid-air
 Power-to-X (e.g. NH3, LOHC etc.)
Storage Economics: Capex, Efficiency…

If roundtrip efficiency is worse,

Cheaper capital and operating costs
charging costs are higher (more
may offset higher charging costs
energy needed)

Lifetime:
Total $
Energy
($/MWh) MWht (discharged)

If this is higher,
discharged energy
must be more valuable

A simple LCOS calculation

ignores other sources of
revenue
(e.g. www.essinc.com/cost-calculator/)
 Capex vs. LCOS
▪ Battery costs scale with additional kWh of storage, other solutions less so
▪ Some batteries (Li-ion) may require major replacement/augmentation costs (e.g. 10 years)
▪ Solutions like PHS may require more space and construction time, but have lower LCOS
▪ As battery costs fall, so will their LCOS!
(and the ‘cross-over’ duration for alternatives to be cheaper will rise)

400MW system, Capex ($m) 400MW system LCOS ($/MWh)

1600 160
1400 140
1200 120
1000 100
800 80
600 60
400 40
200 20
0 0
Li-ion V-redox PHS Li-ion V-redox PHS

4 hour 10 hour 4 hour 10 hour

Data source: Platte River Power Authority (2019)

 Sizing PV + storage (e.g. to mimic gas cc)
Aim: sizing a battery to generate the same electricity at the same time as a gas plant
▪ Typically requires oversizing a PV-plus-storage project

e.g. 1 MW gas vs. 4 MWdc PV + 1 MW, 4 MWh storage

Battery charging from solar, when this exceeds a gas generation profile
Some solar is curtailed and overnight gas can’t be replicated

Source: Bloomberg NEF (2020)

 Sizing PV + storage (e.g. to mimic gas cc)

Increase PV to 7 MWDC and storage to 6 MW, 24 MWh:

▪ More excess solar can be stored
▪ Same round-the-clock output as the gas plant can be delivered
▪ In reality, multiple PV+S sites would be used, not one large one
▪ In less sunny places, a larger PV : storage capacity is needed

Source: Bloomberg NEF (2020)

Model.energy

A simple online model for ‘baseload’ supply, using renewables + storage

 LDES v. Li-ion vs. Hydrogen?

One study, looking at different scenarios (for US, net-zero by 2040)

Source: LDES Council (2021)

 More modelling: Germany example
An example, looking at how to cope with “dunkelfläute”:
▪ “Storage requirements are defined by a 12-week or longer period of intermittent scarcity”
▪ System planning based on average years underestimates storage requirements and system costs

Modelled outcomes
Power:
▪ Solar PV and onshore win: nearly twice as installed capacity as 2020
▪ Offshore wind power: a tenfold increase
▪ Storage discharging power capacity is mostly hydrogen-fired CCGT
▪ Installed CCGT capacity  average load
▪ Total discharging capacity can supply 77% of the peak load
▪ Storage charging power capacity is about three-quarters of discharging capacity
▪ Up to 161 GW of renewable generation is curtailed (cheaper than building more storage)

Energy:
▪ Most storage energy capacity is hydrogen
▪ Storage energy capacity is about 7% of annual load (24 days of average load)
▪ 23 days for hydrogen, 6 days for pumped hydro, 6 hours for batteries
▪ The total primary supply from renewables is roughly 130% of annual load
▪ 65% of primary energy directly serves load, 23% is charged into storage, 12% is curtailed
▪ Storage discharge accounts for 17% of load
Pumped storage growth
 New pumped storage in China
The 3.6GW Fengning Pumped Storage Power Station
▪ Commissioned at the start of 2022, by the State Grid Corporation of China
▪ Located in China’s Hebei province
▪ 425m head height
▪ 12 reversible pump/generating sets of 300MW each
▪ “Designed annual power generation will be 3.4 TWh”
▪ 75% round-trip efficiency
▪ $1.87 billion project,
▪ Two 1.8GW phases, started in 2014
▪ To operate as a peaking power plant to balance power supply from large wind
and solar parks located in northern Hebei and Inner Mongolia
▪ First pumped-hydro storage plant in China connected to a DC network
▪ China is targeting 62GW of operational pumped-hydro facilities by 2025,
120GW by 2030
▪ Currently, China has 30.3GW
New, “Brownfield site” Utility-scale PS

e.g. Kidston Gold Mine in North Queensland.

▪ First in the world to use two disused mine pits for hydropower.
▪ Using existing infrastructure from the mining operation should lower
construction costs.
▪ Paired with 270 MW Solar PV
▪ Requires 187km transmission line
▪ Financial close reached May 2021. Project connected by 2025?
▪ 250 MW, 2000 MWh.
▪ 30s ramp-up to full generation.
▪ Cost ~$780m
Coal-to-storage (Kentucky, US)

▪ 200 MW, 8 hour storage

▪ Developer: Rye Development
▪ Former coal strip mine site
▪ Estimated investment of nearly $1 billion
▪ Filed for a permit and commenced the development process (Jan 2022)
▪ Would receive a 50-year license to operate

“As coal is phased out, communities face uncertain employment & economic
futures. Projects like Lewis Ridge create new jobs and economic activity”
 Compressed Air (CAES)
Not a new technology!
▪ 1991: Alabama Electric Cooperative started operating a 110 MW CAES plant in McIntosh
▪ Project had a 26-hour storage duration
▪ Served to smooth demand between the low weekday loads and high weekend peaks
▪ Acquired by Siemens in 2015
▪ Still the only operational utility-scale CAES plant in the US!
▪ The Los Angeles Department of Water and Power (LADWP), is evaluating proposals for new CAES deployments, amongst others

Characteristics:
• Early examples burned natural gas to avoid cooling
on expansion
• New designs store compression heat and re-use it
(“adiabatic”):
• Lower emissions
• Better round-trip efficiency (up to 70% vs. around 50%)
• Charge/discharge response similar to
pumped hydro
CAES in China

May 2022
▪ Changzhou, east China's Jiangsu Province
▪ Operated by Huaneng Group
▪ The salt cavern is about 1,000 meters underground
▪ Energy storage capacity is 300MWh
▪ Power capacity 60MW
▪ 60% round trip efficiency
 Other CAES examples
Example: HydroStor
▪ Rosamond, California, USA
▪ Capacity/Energy: 500 MW and up to 12 hours
▪ In-service by 2024
▪ Applications: “renewables integration, replacement of fossil fuel
generation, transmission optimization”
▪ “No chemicals or contaminants, no disposal liabilities, no fire risks”
▪ “Smaller land and water footprint compared to pumped hydro”
▪ “50-year lifespan, unlimited cycling, no performance degradation”

Example: Storelectric:
▪ Proposing retro-fit approach (to CCGT):
▪ “reducing emissions by a third and adding storage-related services”
▪ “a minor modification”
▪ Scales from 20MW to multi-GW, durations of 4+ hours.
▪ “Cheaper per MWh than both batteries and pumped hydro”
▪ “Able to be installed closer to both supply and demand than hydro”
▪ “Enough geological capacity in the UK to provide two weeks of back-up”
 Other “Gravity” Storage Innovation?
e.g. Energy Vault
• A 120-meter tall, six-armed crane stands in the middle.
• In the discharged state, concrete blocks weighing 35 tonnes
each are stacked around the crane far below the crane arms.
• To ‘charge’ crane arms locate concrete blocks and power is
used to drive a motor and lift them off the ground.
• The system is “fully charged” when the crane has created a
tower of concrete blocks around it.
• To discharge, blocks are lowered, driving the motor in reverse
to generate power.
Status?
• In August 2019, SoftBank’s Vision Fund agreed to invest $110 million.
• Aim is for 4-8 MW of continuous power for 8-16 hrs and 80-90% efficiency.
• Design life of 30+ years, no degradation.

▪ Korea Zinc subsidiary (Sun Metals) refinery in Queensland

▪ 2nd largest electricity consumer in Queensland (>1 TWh/yr)
▪ Energy Vault to use waste materials for its composite blocks Source: energyvault.com
 Weights in mine shafts??

Status:
• “£1m” 250kW energy storage demonstrator in Scotland
• Trials begin February 2021
• 15-metre high rig, two 25-tonne weights
• 4MW scale-up project due “later in 2021”

Claimed characteristics:
• Full power (in <1s) by dropping both weights
• Output smoothing by using single weights over longer time periods
• Efficiency claimed >85%
• High c-rate possible (8+)
• Small surface footprint: 30m x 30m for 8MW
• No degradation, DoD limits etc.
 Thermal Storage (hot)

Status:
▪ 30MW pilot plant in Hamburg, Germany
▪ Can store 130MWh for 1 week
▪ Site is a decommissioned power plant
▪ 80% of components are off-the-shelf & re-used
▪ System ‘up to 500MWh’ due by 2022

Characteristics:
▪ Converts electricity into hot air using a resistance heater
▪ Heats 1,000 tonnes of volcanic rock to 750°C
▪ Conversion back to electricity uses a steam turbine
▪ Good insulation means heat can be stored for 1 week+
▪ Claimed LCOS of €40 ($48) per MWh
▪ The inventor, Henrik Stiesdal is developing a modular version
for use at wind and solar farms (Stiesdal AS)
Liquid air energy storage (LAES) – cold!

Uses energy to compress and cool air to -196 °C

▪ ~700 litres of ambient air  1 litre of liquid
▪ Stored in insulated tanks at low pressure
▪ For discharge, the liquid air is reheated
▪ Resulting high-pressure gas drives turbines
▪ Waste cold and heat from the process are stored separately

Vecci et al., 2021

 LAES
Liquid air energy storage (LAES)
• Technology based on established turbo, power
generation and industrial gas sector equipment
• 60-75% round-trip efficiency (Nov 2020)

• Using waste cold and/or waste heat Highview Power:

increases efficiency (potentially to 100%) ▪ 50MW / 250MWh LAES facility in Manchester, England.
• Fast response can be added using flywheels or ▪ Due to enter into commercial operation in 2023
supercapacitors: enables <1s response ▪ Revenue through “the Capacity Market, grid balancing,
• High energy density arbitrage and ancillary services such as frequency response
and voltage support”
• Lifespan ‘30-40 years’ ▪ Also consented as being fully ‘Carbon Capture Ready’

Jan 2022:
▪ Partnership between Sumitomo SHI FW (Japan) and the
Shanghai Power Equipment Research Institute
▪ To study the use of LAES (developed by Highview Power),
in 4 and 8-hour configurations
▪ Both systems with power capacity of 50MW Pilsworth demonstration plant, operating since 2018 (UK)
Power-to-’X’ (H2)

Hydrogen & future ‘electrified’ energy systems:

▪ Electrolysis as a tool for increased direct electrification
▪ Flexible power consumption: “use it or lose it” (avoid local curtailment)
▪ Pipes & storage instead of excess power infrastructure (= cheaper system?)
▪ District heating networks could use excess heat from electrolysis
▪ Various industries can use co-produced O2
▪ If insufficient local hydrogen demand: export it!
(Energinet, 2020)
Full electrification (idealised)

Storing renewable power is essential:

Full electrification (with hydrogen)

Storing renewable power or moving it into other sectors?

Long-term storage & large-scale balancing

Green Hydrogen Hub Denmark

▪ Gas Storage Denmark, Eurowind Energy, Corre Energy
▪ 350 MW electrolysis by 2025  1,000 MW by 2030
▪ 200 GWh hydrogen storage (underground salt cavern) by 2025  400 GWh by 2030
▪ 320MW ‘Hydrogen Compressed-Air-Energy-Storage’ (CAES) by 2025
▪ Hydrogen will be supplied to industrial consumers or can be converted back to
electricity to help with security of supply.
 Risks & Complexity…

No two storage business cases are the same!

▪ “Storage” covers diverse applications, localities, scales, revenue streams and

technical solutions: understanding specific project drivers is key!

▪ The biggest values now may not be the biggest ones in future!

▪ Past policies and markets were built around generators or users: storage is both!

▪ Storage without a suitable market/regulatory environment may be cost-effective

at system level, but not at project level!

▪ Storage doesn’t have the market to itself: there are alternatives!

Storage: Some Key Questions to Move Forward

1. Which specific problems and costs can storage remove or reduce?

2. If deployed for #1, what other values could storage bring to the system?
3. Does the market/regulatory environment enable these to be monetised?
4. If not, what market or tariff-setting mechanisms need to change?
5. By who and at what cost can storage by deployed?
6. Based on detailed load mapping, where and how much storage is optimal?
7. What are the resulting cost savings and additional values?
8. Where and to who do these financial benefits accrue?
9. Can costs (#5) be recovered directly through the financial benefits (#7, #8)?
10. Which additional revenues, transfers or cross-subsidies are required?
THANK YOU
Infocus International
+65 6325 0210
enquiry@infocusinternational.com

Do not copy or redistribute without written consent of Infocus International and faculty