Handling the heat: How smarter distributed solar, batteries & load control can support energy system resilience
With record-breaking high temperatures, Australia’s energy system was under stress last month. Temperatures soared over 40C in South Australia and Victoria, driving electricity demand up as homes and businesses cranked their air conditioning units in an attempt to stay comfortable. This pushed electricity demand close to the levels of available generation, which led to extremely high wholesale energy prices as more expensive generation was require in both states, and ‘load shedding’ - i.e. utility initiated power outages - was implemented in Victoria.
The question of why there was a need for load shedding is beyond the scope of this article, but the Australian Energy Market Operator (AEMO) acknowledged (and apologised for) controlled outages that affected 60,000-100,000 Victorian consumers at a time. These were necessary to meet 250 megawatt (MW) supply shortfall during the ‘peak’ afternoon consumption periods; this shortfall would have been more severe had the Reliability and Emergency Reserve Trader (RERT) not been called into action to help balance supply and demand.
Rooftop solar helps ease the pain - but is currently limited
SwitchDin’s technology provides electricity networks and retailers with the tools to manage fleets of distributed energy resources - simply and securely.
Graph via Australian Photovoltaic Institute (APVI)
It would also have been worse without behind-the-meter rooftop solar, owned by hundreds of thousands of homes & businesses across the state. While these homes & businesses managed to reduced the stress on the grid by meeting a portion of their own energy needs locally (as opposed to drawing from the grid), only a tiny fraction of them are equipped with battery storage. Without batteries - or more specifically, a multi-mode inverter/battery setup that can transition the site to ‘off-grid mode’ in the event of a grid outage - solar-equipped sites suffer blackouts in the same way that non-solar sites do: there’s no electricity.
The reasons for anti-islanding in solar are sound: if solar PV systems continue to produce power during an outage, technicians working on the lines could be electrocuted, and equipment on the local network could be damaged. So it is important that systems react safely to outages.
That being said, the phenomenon does leave many solar system owners shaking their heads about the fact that they have an on-site generator that doesn’t work without a live grid - or even more confusingly, having their power turned off despite the fact that they are contributing energy to the system, which seems like an egregious waste of resources. (The frustration level would be even higher for sites with solar and battery storage which are not capable of operating in off-grid mode, as is sometimes the case.)
The controlled outages deployed by AEMO in Victoria are extremely unusual events, but they raise the question of the role that distributed energy resources can/will play in contributing to the future security of the grid. SwitchDin has developed technology to underpin future solutions to this challenge, but other infrastructure - and more importantly, deep collaboration with network operators and AEMO - would be a prerequisite for anything to be implemented.
Here we take a look at what this improved resilience might look like in practice, and how it would tie in with other elements of a future energy system that proactively and constructively integrates distributed energy resources into its operation.
Building resilience at the site level (homes & businesses)
An aside: Why battery-based ‘off-gridding’ isn’t the answer
Enabling a degree of energy self-sufficiency (even if only for short periods) with on-site batteries is the easiest-to-understand approach to building energy resilience for homes & businesses. In its simplest form, it would mean that as many sites as possible are equipped with battery storage with backup functionality and (ideally) an intelligent way to pre-charge those batteries in anticipation of periods of high demand (and possible blackouts).
These individual homes and businesses would a) greatly reduce their reliance on the grid during periods of high demand (as Finn at SolarQuotes did), and b) would be able to keep power on in the event of an outage.
It’s immediately clear, however, that this approach leaves open some serious gaps, especially if it is not augmented with solutions that take into account the ‘bigger picture’ and infrastructure required to support a broader solution. Firstly, not everyone can afford (or has space for, or wants to own) a battery bank - so at most this only one piece in a multi-layered puzzle.
Secondly, the idea of planning for a rare contingency (e.g. unplanned outage or load shedding) by equipping everyone with on-site (currently expensive) batteries is antithetical to the purpose of the current energy system, which is to deliver affordable and reliable electricity supply to everyone. Although stand-alone power systems may work for some regional electricity networks who service regional communities, ‘batteries for all’ would be an extremely expensive and ineffective approach to solving the ‘problem’ of energy security on the National Electricity Market or WA’s Wholesale Energy Market (WEM), for example.
Thirdly, batteries come in a range of sizes, with a range of capabilities and degrees of software intelligence, so not everyone who has one will ride through a blackout in the same way. In a nutshell, relying heavily on on-site batteries to service their local loads (and not provide other services) is an ineffective approach to building system-wide resilience. Let’s leave it to the doomsday preppers.
Small-scale, distributed batteries will be important in the future of our energy system, but the role they play will be multifaceted - doing more than just providing backup power to homes during a power outage. For example, batteries can be deployed as part of a virtual power plant (VPP) to sell into energy markets or provide energy market services like frequency control ancillary response (FCAS). On-site batteries may also be controlled by distribution networks to provide stabilisation support at the local level (like voltage regulation or demand management) or to prevent reverse power flows by managing generation.
But batteries are just one piece of several common pieces of equipment that can be controlled. Others include:
Solar inverters
control for solar generation ramping and curtailment (depending on local network circumstances)
often the de facto controller for a battery bank
off-grid mode (for systems with battery storage)
Other demand response enabled devices (DREDs)
air conditioning units
water heaters
pool pumps
electric vehicles
Diesel, gas micro-turbines and other combustion generators
From SwitchDin’s point of view, managing distributed energy resources (DERs) - which may incorporate any combination of the above items - in the big picture of the energy system starts with visibility. In particular, the ability for a management system to ‘view’ sites in their entirety - including their capabilities and constraints across all their devices and loads in realtime - is crucial for maximising efficiency. This is especially true when VPPs eventually come to incorporate tens or hundreds of thousands of individual sites.
The prerequisite for this level of visibility - which is itself the prerequisite for advanced, smart control - is seamless integration between all of the composite devices in the system. Will all of the devices speaking the same language, it is possible to connect them into a management platform that allows them to be aggregated into a VPP.
This device-to-control-room management capability is where the contribution of DERs to improving energy system resilience begins. Not coincidentally, it’s also where the new relationship between energy prosumers and the electricity system starts to take its shape.
Building resilience at the network level
Distribution network service providers (DNSPs) are at the forefront of the changes afoot with the rise of rooftop solar in Australia, and among the first who must address the challenges associated with it. While rooftop solar has been great for energy end users, high solar penetration levels have led to localised voltage issues on some sections of some networks; another common issue is reverse power flows through transformers (which are generally not designed for two-way traffic).
To date, many of the methods used to deal with these issues have often been blunt: limits on solar inverter sizes, solar production curtailment and blanket zero solar export rules. But as technology (such as SwitchDin’s) designed to offer more granular and subtle control comes into existence, networks are piloting more sophisticated solutions which give customers more choice and networks control of energy flows at critical times. Generally, this approach involves controlling a number of DERs operating ‘under’ a local distribution transformer in a feedback loop with the transformer itself.
An overview of battery use cases for customers (homes & businesses), utilities (e.g. integrated utilities or networks in the Australian context), and energy markets (e.g. as a tool in AEMO’s toolbox). This chart is from a report by the Colorado-based Rocky Mountain Institute with a focus on the US energy system (hence the different terminologies), but is applicable to Australia as well.
In deploying methods to manage resources at the local network level, DNSPs are building the infrastructure and gaining the institutional knowledge necessary to optimise the use of DERs to improve grid resilience going forward. This could involve siting community batteries alongside the local infrastructure, which can be managed in conjunction with other resources (including customer-owned solar & batteries) to strengthen the ability of individual sections of the grid to withstand turbulence, possibly temporarily islanding off into ‘microgrid’ mode when power from the mains is lost.
But a community battery would be an underutilised investment if its sole purpose was backup power. As with household batteries, this sort of grid-scale battery will come into its own when it is possible to unlock multiple value streams, which would include not only network support but also energy market participation.
Building resilience at the energy system level
Which brings us to the ‘big picture’ of Australia’s wholesale energy market(s). SwitchDin has estimated the enormous potential for VPPs to help contribute to demand on the NEM (not to mention the WEM) using batteries alone - potentially up to 11 gigawatts (GW) by 2040 in a high growth scenario. (This number doesn’t include the grid-scale or distribution network-level batteries.)
This would be a substantial amount of visible, dispatchable generation that could be called to respond when energy demand threatens to outstrip supply. But the main purpose of the composite resources would be to directly service the energy needs of their owners; as such, it’s likely that only a portion of this potential 11GW will ever be deployed at once, and even then only a small number of times per year (if at all) - and only if there is a compelling financial case for the system owners to do so. (Local network or ‘community’ batteries could be dispatched in a similar capacity.)
In terms of energy system security, this means that electricity retailers, integrated utilities & AEMO themselves will have another large lever to pull to keep supply and demand in balance without unduly inconveniencing customers (as the response would be distributed across thousands of homes).
To put this in the context of the rolling blackouts that occurred in Victoria: A ‘simple’, battery-based VPP comprised of a rotation of 80,000 homes at a time (e.g. out of as many as 7 million battery-equipped sites by 2040) may have been able to contribute 250MW to absorbe this impact and potentially keep the power on. In this scenario, both the homeowners who provided battery power as well as those whose lights stayed would benefit.
Distributed demand management: It’s all about communication
If you have doubts about whether this is feasible, consider this story from cutting-edge startup electricity retailer Amber Electric. Amber passes through the wholesale spot market electricity price to their customers instead of charging a flat set of rates, as most retailers do. This means that customers benefit when wholesale prices are low, and that they are strongly incentivised to reduce their usage (or export energy) when prices are high.
Amber warns their customers of possible high price events in advance so that they can take action accordingly. In effect, this is a version of relatively low-tech but effective demand management - strategic reduction of loads to balance electricity demand with supply.
In a Twitter thread posted after the heatwave events, Amber CEO Chris Thompson noted that on one of the record high temperature days, half of the company’s customers significantly reduced their consumption (e.g. turning down their AC) in an effort to minimise their exposure to the high prices.
This is an easy-to-understand example of how demand management can work. In a conversation about the heatwave, Business Renewables Centre Australia technical director Jonathan Prendergast noted that the fundamental principle behind all demand response is communicating to end users to lower their consumption to reduce stress on the energy system.
On the most elementary level, this could be as simple as a public service announcement; however, to maximise energy security and reliability (AEMO’s mandate), distributed demand response will be best managed through technical channels. SwitchDin’s technology enables device-to-control-room management of DERs by networks, retailers and system operators across a wide range of products and system configurations.
The enormous opportunity in distributed demand management
Although virtual power plants have come to be synonymous with distributed battery storage (or ‘battery swarms’), direct control of key behind-the-meter appliances such as air conditioners holds even more promise, depending largely on how consumer interest in such programs is stimulated.
At the moment, demand management is primarily a tool for network operators and AEMO (e.g. via the RERT) and going forward its importance in the broader energy system will only grow. In the case of Victoria, for example, a temporary, rotating 2kW reduction in loads (e.g. air conditioning, hot water, pool pumps) across ~100,000 homes may have averted or mitigated the need for load shedding, with minimal impact on the individuals whose resources were deployed. This would have been a preferable outcome to the rolling blackouts that occurred.
While there are quite a few technical, logistical and administrative hurdles to get to reach this point, automated, distributed demand response will certainly play a significant role in Australia’s future energy system.
harmonising VPP performance across all levels
Energy market participation is one of the most readily-understood ways in which DERs can contribute to the energy system. Until now, talk about VPPs has largely revolved around batteries selling energy into the grid during times of high demand / high prices, but as discussed here the opportunity to incorporate more diverse resources to provide more diverse services is larger than this.
As VPPs grow in size and number, so will the potential for their management to get complicated and prone to risk, threatening system stability. One of the biggest challenges as VPPs scale is managing the conflicts involved when DERs are deployed for both network service and energy market purposes. (AEMO is building frameworks to prepare for the growth of VPPs on the NEM by addressing this type of challenge, and SwitchDin is part of a consortium lead developing a tool to address this challenge - more to come on this soon.)
The smallest yet possibly most important element of the harmonisation process in DER management will be the human element: How will homes and businesses feel about having their devices controlled by a gargantuan system that is asking them to contribute to stability when providing stability is written into that system’s mandate for operation? We don’t have those answers, but providing the right incentives for end users to participate willingly will be essential.
Smart infrastructure & management will be key
Distributed energy resources hold massive promise for building energy system resilience at the site, network, and system-wide levels. But integration between the wide range of components used in consumer DERs is a major barrier to building, operating and scaling virtual power plants.