With lithium batteries all the rage and more and more battery companies offering more and more products, how does the customer pick the right system?
Batteries can be confusing: different battery chemistries (LFP, NMC, LTO), sizes, price ranges, required ancillary equipment and installation costs all factor in to the decision. To help out, we’re going to consider the following questions a customer should ask.
1. Is it safe?
2. What is the real cost?
3. Is it the right size?
4. Is it right for what I need?
5. Where does it make economic sense?
Is it Safe?
Safety should be the number one issue when choosing a battery system however, discerning safe from unsafe can be difficult. There are a number of safety certificates and standards the batteries should have: UL9540A and IEC 62619:2017 are probably the main ones. So this is a good place to start. The Clean Energy Council (CEC) is making life a bit easier for this by giving accreditation to those companies that apply and have appropriate certification. But of course, just because a company can get some samples through this sort of testing doesn’t mean they have quality production in place to deliver consistent, quality cells and modules year after year.
An in-depth assessment of battery companies is obviously well beyond the reach of most buyers. However, one basic thing anyone can do is an internet search. It’s easy to check if the supplier is a significant and well-known producer: there are now several hundred battery companies in China with obviously much varied and unknown quality, so if the battery provider is relatively unknown and small, watch out. You can also search for safety incidents involving the battery company you are considering. Sticking with Tier 1 producers is a good potential strategy but, as we’ve found out with incidents involving Samsung and LGChem, even that is not a guarantee, although with LGChem the problem was with larger systems and with Samsung the batteries in Galaxy phones created all the issues. Some companies have installed tens of thousands of home energy storage systems and never had a ‘glitch’.
Another option to consider is whether the company is supplying either to major car companies or the US defence force. A company like Kokam for example, has supplied batteries for submarines so must have passed some of the most stringent testing on earth.
There have been a few companies pushing lithium ion phosphate (LFP) chemistry over nickel manganese cobalt (NMC) chemistry recently and we have discussed this comparison at length in previous articles. While it is true that LFP has inherently less heat generation when problems occur, and is easier to manage, using LFP in itself is not an iron-clad guarantee of safety. Things can still go wrong, particularly with battery management. LFP is easier to get right but good companies have also got NMC ‘right’: it’s just much harder.
The other major component of safety is at system level. Making sure the batteries communicate properly with inverters without compromising the management system, and ensuring that proper mitigation is in place if things go wrong, are also crucial elements. This is obviously more of an issue for larger, customised batteries than the ‘plug-n-play’ products for homeowners.
Recently the government has done a couple of things to make life a bit easier for the average homeowner and industrial/commercial businesses. As mentioned previously, CEC accreditation for batteries has been brought in: this is given to batteries that have the right certifications in place. Also a number of Australian standards have been developed covering both the batteries themselves and their enclosures.
The Clean Energy website is a good reference tool.
So in terms of safety homeowners and small businesses can do a lot of homework via internet sources and check that the systems are either CEC approved or have the right certification. Larger businesses looking at larger installations should have enough money to pay for technical experts to drill deeper into this issue.
What is the Real Cost?
You would think this would be relatively straightforward but it seems with batteries nothing is straightforward!!
First there is the cost of the battery unit. The unit will specify an amount of energy (in kWh). The first question to ask is, is the number that is quoted the nominal energy capacity or the useable energy capacity? Most battery systems are not used to their full energy capacity in order to extend life. For example, the Telsa Powerwall 2 has 14 kWh of total energy but 13.5 kWh of useable energy. This is a very important point to clarify: useable energy can be up to 20% less than total energy.
The next thing to consider is what are the extras needed? For example the Powerwall 2 needs a ‘gateway’. A few systems are ‘all-in-one’ meaning they have an in-built inverter already built in (eg. Alpha ESS). But for most systems, you will need a hybrid inverter (a hybrid inverter sorts out both the solar input and the battery). The good news is that hybrid inverters now aren’t that much more expensive than normal inverters and prices are coming down all the time. SolarQuotes has an excellent comparison of different battery systems including whether they are ‘all-in-one’ or whether they need an inverter.
Extras for buyers of larger systems may also include design and building of the enclosing structure. This is a key component to address for larger systems: what are you actually buying? Just battery modules plus battery management system, or a full, complete turnkey solution? With Australian standards being constantly upgraded, getting a system that complies with relevant Australian standards is becoming a key requirement.
Installation cost is also very important and this can vary greatly. Some systems are well designed with straightforward installation in mind, whereas others are much more complicated. Getting an accurate installation price upfront is important.
Then there’s lifetime. This is a critical issue and one every buyer needs to be wary of. Results from the Australian Battery Test Centre (batterytestcentre.com.au/reports) show just how critical an element this can be with most systems tested, including Powerwall One, yielding relatively poor lifetime. Expecting 10 years out of your battery only to find that you only get five is obviously not ideal.
Warranties will typically have both a time period and a number of cycles or total energy limit. A 10 year warranty and at least 3,650 cycles (one cycle per day for 10 years) is typical for quality systems. The warranty will typically specify a percentage of the original capacity at the end of the warranty. For example, the Powerwall is 70% of initial capacity, others have 80%. Industrial/commercial users will need to carefully consider how they are using their batteries when considering cycle life - another key factor is rate, i.e. how fast a battery is charged and/or discharged (homes are typically slow charge and discharge so rate isn’t a dominant factor).
Most battery companies have options for much longer cycle life cells, eg. up to 20,000 cycles. These are always more expensive and may only be worth it if you are actually using the battery for that many cycles. The majority of homeowners only use their batteries once per day so a warranty for 3,650 cycles or thereabouts is usually more than enough. Note that the batteries aren’t useless at the end of their warranty: they will still have 70-80% of their original capacity. What happens at the end of warranty? That’s a question that needs a whole other article to consider.
So how much should a homeowner expect to pay, and does it make sense? Solarquotes gives a good price comparison, per warranted kWh at one cycle per day, that is very useful for the homeowner (solarquotes.com.au/battery-storage/comparison-table).
Some interesting highlights from the table. Tesla Powerwall 2 is 31c, LGChem Resu 13 is 24c, and Zenaji Aoen is 21c. Pylontech is a battery company doing very well at the Battery Test Centre but is not in the table: they come in at about 23c.
If you look at these numbers you might think great, I’ll either break even or make a little money. However, these numbers don’t include installation costs and extra inverter costs. We’ll dive into some more detail on costs later on.
The Zenaji system is interesting as it is based on a safe, long life chemistry called lithium titanium oxide or ‘LTO’. It has low cost per kWh although you will pay more upfront for this system: it gets down to 21c by spreading initial cost over a 20 year warranty. The customer needs to choose between doing this, or spending less upfront and doing a partial or full replacement at 10 years. Applications doing multiple cycles per day obviously greatly favour the Zenaji system. Zenaji has put a battery into the Battery Test Centre for evaluation and have recently applied for CEC approval, both of which will give their products more impetus.
Another primary factor that can change economics considerably is the cost of capital. This is particularly important when deciding between a higher cost, longer life option and a lower cost, shorter life option. At the moment one could argue that the cost of capital is realtively low (low interest rates). However, larger installations and investments will always have alternative options to invest money.
To illustrate how much this can affect economics, I’ve used two hypothetical batteries (battery cost only) with different costs and different lifetimes and calculated a cost per kWh of total delivered energy.
1. $700 per useable kWh, 10 year lifetime to 80% capacity
2. $1200 per useable kWh, 20 year lifetime to 80% capacity
The table shows the results in cents per kWh delivered energy for different capital costs.
Measuring the capital cost changes economics considerably and also can change the best battery option.
Of course, in many scenarios paying more upfront is a moot point: many investors have a certain payback period they’re comfortable with, so even if the more expensive option makes sense in the longer term, the investor may not want the risk of the associated longer payback period.
Is it Right for What I Need?
This is probably more relevant to industrial and commercial use rather than home use as most homes use storage the same way.
The rate at which the battery is charged and/or discharged greatly impacts on battery life. Batteries that are designed to give similar life at higher rates will be more expensive than low rate batteries. It is critical to understand the use conditions with respect to rate, both to ensure a suitable battery, and to make sure the system is covered under warranty. This is not a concern for homes where batteries are charged and discharged slowly but can be significant matter in businesses particularly with high power equipment.
Temperature is another important consideration especially in some areas in Australia that obviously can get very hot. LTO systems like Zenaji’s have a much better temperature tolerance. Other systems may need an air-conditioned enclosure to ensure proper battery operation. Cold environments have similar issues.
What Size is Right?
Since batteries are expensive this is important. There are a wide range of sizes available to the homeowner, from ~2 kWh to ~14 kWh. Note that 1 kWh is the amount of energy obtained by using a power of 1kW for 1h.
It’s not that hard to estimate a maximum size that makes sense. For those with solar already, if you have a smart meter you can easily see how much energy you are using and how much is being produced by solar.
The diagram shows an example of a house with solar power production in green overlayed on power use.
The total energy available to put into the battery is the green area: this is the excess solar production. So it makes no sense to have a battery that is bigger than this (one obviously needs to look at an average or higher than average day). Then you have to ascertain the total amount of energy you are using from the grid. Again, it makes no sense to have a battery larger than this, even if you have enough excess solar power to charge it.
If the energy you are taking from the grid is more than your excess solar then you could think about installing more panels.
Of course, we are talking maximum battery storage here: you always have the choice of a smaller system if you don’t want to spend that much money.
If you don’t have a smart meter that gives this level of information you can get a general idea from your power bill.
The excess solar for the quarter is how much you fed into the grid. Also, the total amount of energy you took from the grid will be on the bill.
Here’s an example.
Solar fed into grid in quarter 1000 kWh
Energy taken from grid in quarter 700 kWh.
So in this case, there is plenty of solar available to offset grid use. But since we’re only taking 700 kWh from the grid, that is the amount of energy we want to put into a battery. To calculate the battery size, we need to spread the 700 kWh for the quarter over ~90 days in the quarter. This gives 7.8 kWh per day. So a battery around 7.8 kWh would make sense.
Here’s another example.
Solar fed into grid in quarter 600 kWh
Energy taken from grid in quarter 1000 kWh.
In this scenario, the customer could choose to stay with the panels he/she has, and use the 600 kWh to partially offset the 1000 kWh. Again over 90 days gives a battery about 6.7 kWh. Or the customer could choose to increase their panels. In this case, they should probably wait for another power bill to calculate their battery size as the energy taken from the grid will change, or get an expert to do some calculations to get a good prediction of what’s going to happen with the new panels.
Larger industrial and commercial users usually need extensive and detailed modelling to optimise storage as their usage and billing can be very different to home and are site-specific. One common scenario is that businesses use high power for short periods of time which causes high demand charges. Batteries can help with this – again, the fine details of the electricity deal need to be closely scrutinised. For example, if the demand charge is based on the highest power use in a quarter, it’s wise to have a back-up generator in case there are a few rainy days where the battery isn’t fully available. The size of the battery is calculated by looking at how much power you need to supply over what period of time in order to drop peak power use to an acceptable level.
Where Does it Make Economic Sense?
For home use, as an example, I’ll use a lower cost battery with a 10 year warranty to 80% capacity - this is close to a best case scenario for residential installations.
For an approximate 13 kWh system, allowing $500 for installation and extra $1100 for a hybrid inverter, 10 years to 80% capacity, and zero capital cost, the cost of this stored energy is approximately 28 cents per kWh. So for most states in Australia (notably not SA), this is higher than simply taking power from the grid. There needs to be another reason to buy a battery (in the absence of government assistance). Obviously, these alternative reasons are highly compelling - helping the environment, power resilience and independence etc.
Probably the best way that prices are going to become realistic and make sense is by extending lifetime warranties whilst maintaining price. Even going from 10 to 12 years moves the cost from 28 cents to 23 cents per kWh which is about break-even in most states. Going to 15 years brings it down to 18 cents, making it exceedingly viable. The technology for achieving this already exists even in conventional lithium-ion batteries. At this point it costs about 20% more to move from a 10 year to a 15 year life however, some battery companies are reticent to reflect the extra lifetime properly in warranties. This needs to change. Hopefully in the next couple of years, 15 year warranties become the norm and open the floodgates for home energy storage.
For industrial/commercial systems in the multiple 10s of kWh size range, costs do not come down a whole lot and larger businesses pay less for power so it can be even more difficult to make batteries work. Where batteries can make sense is where:
(i) The company is paying high demand charges that can be mitigated with a battery;
(ii) The company is off-grid and solar/battery is competing with diesel generation.
Again, extending warranties to 15 years will transform this situation.
For even larger systems, i.e. tens of MWh, the position changes significantly. At this scale, market prices for battery modules plus BMS are around USD 180-200 per kWh. Full systems can be purchased between about 450 USD per kWh and 1100 USD per kWh depending on use, quality etc. These scenarios need careful and diligent quantification on how the battery will be used: system design and implementation is critical. Making batteries of this size safe is obviously paramount: they can go very badly wrong. Companies looking at systems this size definitely need to engage high level expertise.
The primary circumstances where big batteries have proven extremely successful (eg. The Tesla battery at Hornsdale) is where they are used mainly in delivering frequency controlled ancillary services (FCAS). FCAS revenue is considerably higher than the normal wholesale market revenue and battery technology is well matched to this role given its fast response times.
It is in this space where big batteries are making their mark around the world.
Unfortunately there is only so much demand for FCAS: at some point this will be met and new installations will not be able to access this revenue.
A big battery can also work well in new urban developments. In these, incorporating a battery can remove the need for an expensive transformer/substation thereby making the battery economical.
So, in summary, there are many things to consider when buying a battery! For the homeowner, there is more and more information available to help on the internet, websites like Solarquotes and CEC accreditation are useful guides and parameters. Homeowners can also do themselves a favour by doing their homework and researching installation costs, what extra equipment is needed, and making sure the warranty is solid.
For larger business and commercial projects: get an expert! There’s too much money and safety risk involved to do anything else.