Energy Transition #5: Energy Storage Systems for Microgrids

As we read previously, Energy Storage Systems are very important for the Energy Transition – most Renewables have “intermittency” which is interruptions to continuous energy generation due to various factors:

  • Solar PV Power – adversely affected by clouds, precipitation, dust – anything in the atmosphere that reduces the solar radiation which activates the PV panels to produce electricity – and of course, the most obvious source of solar interruption is night time;
  • Wind Turbine Power – meteorological changes to the prevailing winds which reduce the directionality and intensity of wind turbines – common industrial wind turbines require average persistent wind (cut-in) speeds >12.6 kph (3.5 m/sec), with 36-54 kph (10-15 m/sec) producing maximum generation power – and for safety reasons, wind speeds great than 90 kph (25 m/sec) require the turbine to be stopped or braked (cut-out or furling speed) – varying wind speeds during an average day result in wind turbines typically operating 70-85% of the time and at approximately 30-40% of their nominal capacity over a year.  There are some large wind turbines configured for Low Wind Speed sites (average wind <7.5 m/sec) which are International Electrotechnical Commission (IEC) turbine classification system IEC Class III – and lower wind speed ratings help by increasing suitable locations and managing electrical storage requirements due to less intermittency.

For Wind power alone, there can be daily variability as shown in this example figure on the left over a one week period (Wind power supply is shown in Green).  Daily load (total demand) is shown with normal cyclical variation in Yellow.  That leaves the daily supply from conventional power generation (net load) as shown in Orange

Challenges come with short peaks of conventional power generation (need to turn on quickly and rapidly provide the power needed) and lower turndown (below what sometimes may be economical or practical).  Clearly there would be an economic balance when sizing the Wind power facilities, maybe diversifying into an additional Renewable like Solar, and most importantly to add the capacity of Energy Storage to smooth peaks and reduce the amount of conventional power generation required.

Even for more persistent Renewables such as geothermal, wave, or biomass, there can be a mismatch between supply and demand curves.  Economic or physical requirements may mean that installed capacity is less than peak demand, but with demand varying there would often be times of excess capacity which could be diverted into Energy Storage to help meet the higher demand periods.

Challenge of Renewable Wind Power Systems

With intermittency an issue for Renewable Wind, it makes sense to have multiple Renewables linked to increased size and duration Energy Storage Systems.  June 2020 in SE Australia was a good example of the challenges for all Renewable energy production. [1]

They report monthly Wind Power energy production systems:[2]

The percentage of installed capacity of the grid connected Wind Power system in SE Australia showed low points lasting for 33 hours on the 5th-6th, 18 hours on the 11th, 16 hours on the 17th, 14 hours on the 26th, 11 hours on the 27th and 9 hours on the 28th. There were other lows of shorter duration.  During these reduced wind periods, the AEMO grid had to rely on conventional power generation (predominantly), Solar PV, and Hydroelectric.

Daily intermittency is shown on this same website for Solar PV power in SE Australia – the daily capacity factor (%) for Solar Energy Production during June 2020 varied usually between 40-50% with a few days below 40% and a few days slightly above 50%.[3]  Solar PV energy was not enough to make up for Wind intermittency (especially with some alignment in the intermittency due to weather patterns).

For 2019/2020, AEMO Wind Power is about 8% of total electricity generation, Grid Solar PV is about 3% and Hydroelectric is about 7.5%.  Behind the meter (residential and commercial) Solar is about 1%.  BESS is 0.4% which appears significantly too low.

AEMO Renewables are therefore <20% of the total power – this means there is a long way to go for their Energy Transition goals.  To support reliable power and grid stability as the use of Renewables grows, it is likely that SE Australia will need to add significant amounts of BESS / ESS (by orders of magnitude).

With the intermittency of Wind Power in certain seasons (and the daily intermittency of Solar Radiation), the use of Grid Solar PV and BESS/ESS needs to be significantly expanded in this region.

It is also likely that Behind the Meter (residential and commercial) Solar will continue to expand significantly in addition to linked BESS/ESS.  The use of local Hybrid Microgrids is likely to significantly increase as seen in other regions of the world with similar intermittency challenges and grid stability challenges.  California is a good example in August 2020 with main grid curtailments even with Conventional Power Generation running at full capacity.  Communities and businesses with the foresight to have installed Hybrid Microgrids are likely happy with their decisions.  Regulatory obstacles need to be resolved however to allow more distributed power systems including increased Renewables and Energy Storage.

Types of Energy Storage

Historically Energy Storage Systems (ESS) involved Electrochemical battery systems, often with conventional lead-acid batteries, even conventional type vehicle batteries.  The life cycle costs of these battery systems were sensitive to depth of discharge and charging patterns.  There were also very good stored energy systems like (1) Pumped hydro storage (using electricity in low demand periods to pump water to a higher elevation where it can be drained down later in high demand periods to power generators) but they needed certain topographical features not available everywhere; and (2) Compressed air energy (where compressors store the air in underground caverns) but they needed certain geological features not available everywhere.  There are a wide range of Kinetic energy and Potential energy Storage Systems[4] but this post will concentrate on Electrochemical and Chemical technologies due to their ability to be located almost anywhere.

Electrochemical battery storage systems have been undergoing significant development over recent years, expanding from Lead Acid into Lithium ion and Redox flow technologies.  Chemical storage technologies have included Hydrogen being produced with surplus power then available subsequently in peak energy demand periods using fuel cells.  There are significant technical and commercial variations between these ESS which can be reviewed.

The duration of energy storage is an important consideration.  Shorter duration ESS technologies like Lithium ion batteries are well suited to intra-day variation (as described at the start of this article) of Renewable energy sources.  Longer duration ESS technologies are also required for extended intermittency of Solar or Wind energies over several days.  Extreme weather events can also disrupt reliable energy supply without increased energy storage.  The US Department of Energy (DOE) Advanced Research Projects Agency – Energy (ARPA-E) has been supporting the development of Long-Duration Electricity Storage (LDES) solutions for 10 to 100 hours of energy storage.[5]

It is also important to consider the time responsive nature of Energy Sources and ESS.  Sudden mismatches between supply and demand (faults) can adversely affect system stability.  Renewables alone cannot provide the stabilising response where the majority of energy supply is from intermittent Solar or Wind.  Inertia is the response provided in fractions of a second to any imbalance.  Conventional power generation with synchronous generators respond automatically and immediately by slowing down, releasing energy stored by the rotating mass inside the generators – this is called inertial response with short power increases of 7-14% within 0.05 seconds of an event and would have a duration of a few seconds.  This means that there is typically something needed called spinning reserve which can be expensive in capital, operating costs, and emissions.

The Gold Fields Agnew Gold Mine microgrid system (previous post) was running reliably at 54% Renewables but had to have two conventional diesel generators running at half load with only some synthetic inertia provided by batteries to meet the system’s stability requirements.[6]  The initial solution identified to reduce conventional diesel fuel costs and GHG emissions was to add more Renewables and upgrade energy storage with more synthetic inertia capability.  Batteries can respond with additional energy supply as fast as the faults can be measured, with reaction times approaching 0.1 seconds which is slightly slower than synchronous generators, but once detected, they can respond dynamically with high ramp rates (much faster than standby generators can be adjusted by governors to push the frequency back up).  Batteries can deliver full output in less than 0.2 seconds and the output can be sustained for a length of time depending on the size of the batteries.[7]  Fast Frequency Response (FFR) is used to enable batteries to provide fault response quickly.

ESS Use Cases

In order to help decide which type of Energy Storage System, there would need to be a review of the potential use cases.  Lazard provides an annual Battery Energy Storage Systems (BESS) evaluation using one set of assumptions:[8]

It would have been more useful if Hydrogen production, storage, and fuel cell systems had been included in this evaluation.  Hydrogen technologies and costs have been advancing quickly and use for an ESS is increasing so the subject will be addressed further below in this article.

Operational Parameters for the assumed Lazard use cases were as follows:

It is immediately clear that these assumptions above are for short-term energy storage durations (in-day intermittency) with storage durations of 1 to 6 hours.  An ESS rated at 5 MWh and 10 MW would be able to supply 30 minutes of power at peak output.  These assumptions are not reflective of the future many foresee.

Examples of installed Solar PV / BESS microgrid projects described in the last article show much higher amounts of energy storage with corresponding increases in time durations (to help minimise conventional power generation costs and GHG emissions) and grid stability (for unconnected “island” microgrids).

Lazard then reported Unsubsidised Levelized Cost of Storage (LCOS) ($/MWh) for the various use cases and assumptions:[9]

Other cost projections[10] published last year favour Li-ion batteries but show a significant future increase in Hydrogen storage (figure below).  Two of the Energy Storage Systems in this figure are not generally applicable due to topography (Pumped Hydro Energy Storage PHES) and geology (Compressed Air Energy Storage CAES) constraints, so we can concentrate on various types of Batteries and Hydrogen at projected LCOS range of $150-190/MWh.

As the Renewable Energy (Market) Penetration increases, the challenges of intermittency and grid stability mean that longer duration ESS will be required.[11]  For context it is good to remember that in the USA and EU, natural gas is stored in quantities equivalent to thousands of hours of consumption and coal fired power plants typically store 30-60 days of coal.  Conventional power generation therefore has long duration storage and large numbers of dispatchable generators to ensure a high reliability electricity supply.  So it seems intuitive that Renewables could also have significant long-duration energy storage and good grid control systems.

As Renewable Energy Penetration increases, storage durations will increase from 1-4 hours up to 10-100 hours or more. 

Not all types of energy storage are suitable for longer duration storage due to their inability to hold charge for so long (e.g. self-discharge due to internal chemical reactions, which reduces the amount of energy available for work discharge).

Battery Energy Storage Systems (BESS)

A summary of some of these battery technologies is included below for general background knowledge.[12]

Lead Acid – They consist of two lead electrodes submerged in a liquid sulfuric acid electrolyte. Technology improvements  have included incorporating a gel or solid absorbed glass mat electrolyte instead of the standard liquid to improve safety.  They are very common vehicle batteries which have been used in certain Renewable energy storage applications even though they have cost issues related to depth of discharge and life duration.

Nickel – They  are similar to lithium-based batteries in their construction, but the use of nickel allows for different charging properties which suit distinct applications. Nickel cadmium and nickel metal hydride (NiMH) batteries provide improved energy and power compared to lead acid batteries and operate in a wider variety of temperature conditions and levels of discharge.

Lithium ion – Very common and popular battery type right now for certain applications.  Li-ion batteries have lithium compound electrodes and electrolyte structure with some similarity to alkaline batteries with the key difference being lighter and significantly more energy-dense than their alkaline counterparts.  Lithium is a relatively abundant element but other associated minerals like cobalt are supply constrained and sourced from unstable locations.  Li-ion batteries can overheat and are limited in durability, charging cycles, and depths of discharge which can impact their performance.  Long-duration storage is a challenge for Li-ion batteries.

Redox Flow – They have  two circulating electrolyte fluids which exchange electrons directly across a shared membrane. These batteries are well-suited to grid-scale storage due to their relatively low energy density and power output. While some power is required for the operation of mechanical components, the battery itself has a low self-discharge rate and can increase scale simply by adding electrolyte volume. Significant investments in space and equipment are required to operate these batteries.  Vanadium redox batteries can be very expensive and much cheaper acidic solution materials like Iron Sulphate + Anthraquinone Disulphonic Acid are being developed.[13]  Unlike Li-ion batteries, flow batteries have independently scalable energy and power performance characteristics.

Second Life EV – The significant projected increase in future numbers of electrical vehicles (EV) means that there will be an increasing number of batteries that will over time come to the end of their useful life for a vehicle (about 65-80% of their initial capacity remaining, able to deliver an additional 5-8 years of service in a stationary application, but whose ability to retain and rapidly discharge electricity is no longer applicable for moving vehicles).  Typically Li-ion, the remaining life of second-life (SL) EV batteries will vary somewhat due to their history (i.e. how many times charged and discharged, the depths of discharge, and thermal conditions) but they have been identified as applicable for stationary BESS.[14]  The Renewables industry (and Circular Economy) needs these valuable batteries and they may especially be suitable for Distributed Energy Storage associated with Microgrids and for Behind the Meter applications (residential and business).  The attractive cost and savings of repurposed systems would be significantly helped by EV battery suppliers if they considered this subsequent application (so that any disassembly is precluded and ability to easily discharge, test, and recharge externally is included in the original design specifications).  Successful trial applications have included:  (1) Nissan Europe Paris office with 12# SL Nissan Leaf batteries for total energy storage capacity of 192 kWh and power capacity of 144 kW; (2) Johan Cruijff Arena Amsterdam with 148# SL Nissan Leaf batteries for a total energy storage capacity of 2.8 MWh and a power capacity of 3 MW; and a large scale installation (3) Lunen, Germany with 1000# BMW i3 battery packs (~90% SL) for a total energy storage capacity of 13 MWh.[15]  Regulatory policies (e.g. the Battery Directive) need to be updated to facilitate this attractive repurposing.

Battery technologies vary in performance(and suitability) in all the parameters – the figure below shows Discharge Time at Rated Power (higher the better for long-duration storage systems) against System Power Ratings: [16]

Energy Storage Systems (ESS)

Hydrogen has several production routes (i.e. Brown, Grey, Blue, Turquoise, and Green) covered in a previous article.  Assuming a means of hydrogen production is used, there are a couple of hydrogen storage alternatives easily available.  Compressing the hydrogen and storage in conventional high pressure cylinders is possible (~600-1000 barg).  An alternate form of storage is the use of containers filled with metal hydrides (~10-40 barg) which can store the same volume of hydrogen as a high pressure tank (at the same tank size), but at a much lower (safer) pressure.  The stored hydrogen is then available whenever is needed to produce energy through a hydrogen fuel cell.

The storage volume and duration are only a function of the tank sizes or numbers and there is no cyclic degradation of the system (like batteries) so hydrogen appears extremely well suited for long-duration storage (as shown in the figure above).

Scalable hydrogen energy storage also helps the transition to a more diverse Hydrogen economy with other potential hydrogen users being able to make the switchover from conventional fuels or energy sources.  More storage can be added as needed.  Achieving long-duration storage of energy up to (and beyond) 100 hours appears very achievable with hydrogen systems.  Production of hydrogen using Renewable energy (Solar PV or Wind) is Green Hydrogen.

In the US Gulf Coast, Air Liquide commissioned one of the largest hydrogen storage facilities in the world in an underground salt cavern (1,500m deep by 70m diameter) in 2014.  This facility is able to store 30 days of hydrogen (~580,000m3 at up to 200 barg, limited by fracture gradient of rock geology) to back-up a large steam methane reformer (SMR) for industrial hydrogen production.[17]  As discussed elsewhere in this article, unique geological storage conditions are not necessarily widespread, so this solution is somewhat limited, but it is being studied for depleted oil & gas fields, salt formations, and depleted aquifers which do exist in more numerous locations.  The same technology is used for underground natural gas storage and underground Compressed Air Energy Storage.

Hybrid Systems

As the cost of Energy Storage reduces and the availability (market penetration) of Solar and Wind Generation increases, power plant developers are combining projects with on-site Energy Storage to reduce the intermittency of Renewable power supply to the grids.  Significant commercial advantages then exist for the developer by being able to guarantee higher reliability for his own energy deliveries as well as being able to support any grid instability issues (for a separate tariff).  Siemens Gamesa in La Muela, Spain has a 2 MW hybrid facility (shown below) with Wind (850 kW Gamesa G52 turbine), Solar PV (816 panels, 245kWp), Conventional Power Generators (3 x 222 kW diesel MTU Series 1600), and BESS (two systems:  (1) 449 kW / 500 kWh Li-ion batteries and (2) 120 kW / 400 kWh HydraRedox Vanadium Redox batteries (shown below)).  A hybrid controller coordinates energy generation and storage to meet the load demands and reduce the energy costs by maximising the integration of Renewable energy.[18]

This integration behind a point of common intersection with the grid is also effectively the same as an “island” microgrid.  Modifying bi-directional inverters for energy storage, Gamesa Electric were able to provide “black start” and “grid forming” capabilities to enable ZDO (Zero Diesel Operation) mode.[19]  This size facility is nominally suitable to cover the residential needs of up to 800 families.

With lower Levelized Cost of Electricity (LCoE), reliability, and grid stability requirements, it is likely that Hybrid Systems with multiple Renewables (plus some Conventional Power Generation for now) are going to be utilised more and more, together with long-duration Energy Storage Systems[20], so it is good to see successful combinations of these technologies with sophisticated digital transformation control systems keeping the integration seamless.  These solutions are part of the Energy Transition.


[1] https://anero.id/energy/data

[2] https://anero.id/energy/wind-energy/2020/june

[3] https://anero.id/energy/solar-energy/2020/june

[4] https://www2.deloitte.com/content/dam/Deloitte/no/Documents/energy-resources/energy-storage-tracking-technologies-transform-power-sector.pdf

[5] https://arpa-e.energy.gov/?q=news-item/why-long-duration-energy-storage-matters

[6] https://reneweconomy.com.au/global-gold-miner-sets-sights-on-realistic-99-renewables-share-on-mining-projects-68499/

[7] https://info.fluenceenergy.com/hubfs/Collateral/Everoze%20-%20Batteries%20Beyond%20the%20Spin.pdf

[8] https://www.lazard.com/media/451087/lazards-levelized-cost-of-storage-version-50-vf.pdf

[9] https://www.lazard.com/media/451087/lazards-levelized-cost-of-storage-version-50-vf.pdf

[10] https://doi.org/10.1016/j.joule.2018.12.008

[11] https://www.sciencedirect.com/science/article/pii/S2542435119305392

[12] http://energy.mit.edu/wp-content/uploads/2018/04/MITEI-WP-2018-04.pdf

[13] https://iopscience.iop.org/article/10.1149/1945-7111/ab84f8

[14] https://cdn.blog.ucsusa.org/wp-content/uploads/AG-Reuse-Brief_5-12-F.pdf

[15] ibid

[16] https://pg.lyellcollection.org/content/petgeo/25/4/501.full.pdf

[17] https://en.media.airliquide.com/news/usa-air-liquide-operates-the-worlds-largest-hydrogen-storage-facility-edf8-56033.html

[18] https://www.siemensgamesa.com/en-int/newsroom/2018/05/innovation-storage-technology

[19] https://www.gamesaelectric.com/gamesa-electric-at-siemens-gamesas-la-plana-hybrid-pilot-plant/

[20] https://hybridpowersystems.org/crete2019/wp-content/uploads/sites/13/2020/03/3A_1_HYB19_063_paper_Klonari_Vasiliki.pdf

Leave a comment

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.