The questionable economics of "Community" batteries

Like many regions of the planet that have the advantage of glorious amounts of sunshine, the uptake of domestic solar has been immense across Australia.

The Australian Energy Regulator’s (AER) State of the Energy Market report has found, in the financial year (FY) 2023-24, rooftop solar electricity generation exceeded 20 gigawatts in the National Electricity Market (NEM).

This was a quarter of the maximum electricity that can be produced in the grid, reflecting an increase of 2.9 GW from the previous year.

By the end of 2023–24, total generation capacity in the NEM measured 81,082 MW and rooftop solar was the highest capacity at 20,159 MW or 25% of registered capacity, followed by black coal at 20%.

But , as I have mentioned previously, the transmission network was architected in a different era and the introduction of massive-scale domestic solar has been steadily impeding on traditional power sources and manifesting as something affectionally known as the 'duck curve'. ( or 'Nessie curve' if your Scottish). 

The Duck Curve refers to a graphical representation of electricity demand from the energy system on days when solar energy production is high and demand in the grid is low during the middle of the day, and when demand peaks in the evening. When plotted on a graph the lines and curves form a distinctly duck-like shape.

Essentially, the Duck Curve represents the potential for energy system instability, as the energy system attempts to cope with extreme changes in demand across different parts of the day.

As more solar energy is exported to the grid, usually across the middle part of the day when the sun is shining, the curves deepen

So much solar energy is generated in the middle of the day on some grids that supply far exceeds demand, driving down the price of wholesale electricity , often into negative territory. 

Generators then have to cope with steep ramp-ups and ramp-downs to accurately meet the electricity demand, which is commonly the case in the evening peak. It turns out while people were busily augmenting their houses with PV solar panels they were also adding air-conditioners!

This has become a significant challenge for grid generators and operators as the need to rapidly ramp up power becomes more intense every year.  Many elements of the grid infrastructure were designed for traditional synchronous generators and a smooth demand curve. 

As the sun goes down, the lights come on and everyone returns home from work to turn on all their appliances there is a multi-gigawatt step change in the network demand in a very short period of time. 

The extreme demand curve rise is problematic for network infrastructure but in addition it has become uneconomical for traditional synchronous generators that cannot ramp up quickly to meet demand. They cannot switch on and switch off their plant quickly so they have to continue operations throughout the day regardless of whether they can make money, simply so they are available for the evening peak. 

All things being equal you would have thought that these were all positive outcomes, cheaper electricity, and economic dynamics that push out old plant.  However this is really an unplanned transition with the current network not in a state to cope with the rapid rate of change. As I spoken about previously , these are all part of the hidden costs of the energy transition.

The ENA claims there is $16 billion in network infrastructure investment required without some form of non-transmission solution, but this seems like a huge underestimation of the costs of soaking up gigawatts of solar at grid-scale given what we are currently seeing in grid-scale storage projects. 

Realistically I suspect you could easily double that in turns of dollars , and that doesn't even take into account the social-license risks in these large infrastructure projects. Even if they do make economic sense it still doesn't mean they are deliverable. I don't see a grid solution coming for this issue as they are already struggling just to commission enough storage for commercial-scale VRE.

There is a feedback loop created by these impacts on the supply side, as represented below, that continues to steadily make the problem worse.

The problems in the network are getting so bad that there are growing calls for network providers to be able to turn off (curtail off) domestic PV en masse and we have already seen the introduction of negative feed-in tariffs to persuade the end consumer to stop feeding the grid, not that they can in may cases.

Feedback from across Australia , as well as published market research, suggests large numbers of  households are keen to install a home energy storage systems and that households find the idea of being largely self-sufficient for power highly appealing.

But this was all true some years ago ,  and legislative support has been tried but has had a marginal impact. Domestic battery installs are still a tiny proportion of PV install base.

Domestic battery installations as not cheap. A Tesla Powerwall 3 is effectively is a $15k+ investment. There are of course cheaper options but currently the ROI is seen to be around 12-15 years and only very recently has there even been a battery on the market with a warranty that extends that far. The prices have stayed relatively static for a number of years, and the other issue is that

.. on average , according to available data Australians move house roughly every 5 years, with over 40% of households reporting a move within the last five years.

Domestic batteries are also excessively expensive compared to their vehicle equivalents. For example a Telsa Powerwall 3 battery is around $1000/kWh, while the battery in a rear wheel drive Tesla model 3 car is around the same price, but only if you assume the rest of the car is free.

Home battery prices don't have to withstand a car crash, or rapid charging, they have simple passive cooling and come in a pretty ugly plastic box. Yes some of them have a built-in invertors and some other control logic, but they are hardly on-par with a semi-autonomous vehicle.

It all comes down to economies of scale and the tipping of the market where volume drives down per-unit costs so that business overheads start to have less of an impact on each battery sold.  Once that occurs the barrier for new entrants starts to disappear and then competition and market innovation continually drive down costs. 

But we aren't there yet, even with the desires of customers to install these solutions it would require a considerable grant , say $10,000 per household to make this a realisable option for most. At around $1000/kWh in taxpayer grants that seems like an expensive solution.

What about V2G/V2H ?

V2G, or vehicle-to-grid , is the concept of charging electric vehicles during the day and then utilising them to feed the grid at night time peaks.  The technology is in its infancy in Australia , but does fit the profile use case as long as people are incentivised to use their vehicles in a way that is grid-supportive (there are usage profiles that would actually make the duck curve worse). 

The issue is that V2G doesn't really seem to be a winner for consumers.

Australians, whether they are individuals who own an EV or fleet operators, think V2G is great for someone else  

And it's the same in many other countries too. Concerns of trust with electricity companies and the degradation of freedom of usage, and the vehicle itself, are major issues. This may change over time, and could obviously be adjusted through incentivisation but given the consumer feedback I don't see end customers are going to be satisfied with this type of NSP-led solution that also degrades their freedom of use.

Direct Vehicle to Home (V2H) seems like a much more likely winner from that perspective, but it is currently non-existent in Australia as a use case to actually power an entire home.

The major issue with all the V2X solutions is the lack of end user control and the fact that your EV isn't usually at home during the preferred "charge-in" period. Consumers are therefore counting on a network provider to manage their vehicle on a daily basis. Not only isn't there a mature enough charging network in Australia to support this, end consumer see self-sufficiency as a target, and this is not that.

So domestic battery solutions seem expensive and consumer don't seem to like vehicle-based solutions. Surely mid-sized "community" batteries , one that can be shared amongst a group of PV-enabled households will provide the economies of scale to bring down the cost.

Unfortunately, at least in the Australian context, there is little evidence that this is the case.

DNSPs see behind the meter solutions as a threat to their business models. Self reliance of consumers en masse is a direct threat to their profits. DNSPs have worked hard behind the scenes in an attempt to lock-out competition and convince politicians that their "community" batteries are the best solution to the problem. 

Not only have they managed to convince the AER to put in a rule-change to allow them access to a market they would otherwise not be allowed to touch due to their monopoly control, they have also manage to direct tax payers money directly to their own projects, circumventing the end customer.

$1000/kWh ? But isn't that on par with the subsidy that would provide a consumer-led solution ?

But it gets worse.

It turns out that the economies of scale of community batteries don't  provide enough offset to overcome the necessary network augmentation and loss factors to support them.  Although the information a little difficult to find , for obvious reasons, when you do look at the final project costs you realise the equation is even worse.

Between 2021 and 2023, 40 batteries will be installed across Melbourne’s east, south east and the Mornington Peninsula as part of an $11 million program – funded with $7 million from us and $4 million from the Australian Renewable Energy Agency (ARENA).

....  

Each of the 30kW batteries has the capacity to service local homes and businesses with up to two hours of energy (66kWh)

That's $4000+/kWh in total spend.  Maybe that's a one off ... Nope

Energy Queensland will deploy 69 batteries with a total capacity of (at minimum) 4,410 kW / 7,470 kWh (24 x 90 kW / 180 kWh ground-mounted and 45 x 50 kW / 70 kWh pole mounted, or equivalent) across Brisbane and regional south-east Queensland.

..

Total project cost $20.35m , $14.03m in tax payer grants.

$2,724/kWh in total, $1,874/kWh in tax payer funded grants , and that's assuming it doesn't run over budget.

So these solutions are easily double the price of a customer solution, have (re-) embedded the monopoly power of incumbent DNSPs and are in no way in-line with consumers desires to be self-reliant in their energy outcomes.

It turns out there is nothing "community" about community batteries, it looks like they are just another taxpayer rort for big business. 

A grant directed at end consumer to install their own solution seems to make more economic sense and perversely is a more of a "community" solution.