By Phil Kreveld
Long, long ago, along with little foot-warming radiators, battery-powered uninterruptible power supplies hid under office desks. They were toys compared to online UPS and diesel-backed UPS (DRUPS) systems—inseparable parts in data centres, hospitals and surgical suites and other industrial and commercial applications.
Related article: Politics and engineering—a dangerous mix
Is the average suburban and regional centre household PV system (and battery) the equivalent of the computer desk denizen in days of yore? The household sector and commercial and industrial PV owners are encouraged to move to battery investment but it raises the question of why. The distribution networks do not allow reverse power flow, so it is either stored in the battery or if there is no storage capacity left, inverters are externally controlled to reduce power output. Allowing reverse power flow invites major investment—an extremely unattractive proposition.
Some solar systems not only have an ‘in line’ battery between the roof panels and the inverter; they also can power domestic loads when there is a grid failure, just like the under-the-desk UPS. Other systems may or may not depending on the scheme adopted under AS/NZ 4777-2-2024 draft standard. Whatever scheme is adopted, anti-islanding of the solar inverter is required as a safety measure to prevent ‘back feed’ and electrocution of linesmen working on power restoration.
Look at the functional diagram of an in-line UPS (not your ancient under-the-desk one). The battery is always being trickle-charged. If power failure occurs, the static switches switches open, and the battery-inverter within a few milliseconds powers the load. When grid power is restored, the static switches reconnect the load within a few milliseconds to the distribution network. The change-over is virtually seamless. Wouldn’t be nice if things could be organised so that the same could be done for Australia’s suburbs? And what if such a giant UPS, could switch off from the grid—voluntarily, because collectively figured, it would be cheaper to run things ‘in house’?
Running things ‘in house’ as the PV and battery installations grow and grow seems like a possibility—but not quite yet. In fact, there is no engineering effort being made in that direction. When it was raised with AEMO, on several occasions, the question was left hanging. Australian Energy Regulator chair Clare Savage has pointed out that home based solar is matching large capacity renewables—and that distribution utilities are facing serious investment in network assets to ‘handle’ the millions of rooftop generators.
Former Clean Energy Council CEO Matthew Warren has gone public about the expanding home-grown capacity clashing with large-scale, transmission grid-connected generation, putting downward pressure on its earning capacity. You’d think that something would have to yield. But not likely, because the Commonwealth now has a capacity scheme, the CIS, which basically underpins large scale investment with contracts for difference, virtually guaranteeing the restitution of generation losses.
How far could the UPS idea be pushed? Here is a reality check: The big generators connected to the transmission grid have individual and independent controls—not so, those millions of solar systems. They do not have any real controls—that is to say, in the hands of their owners. Any control available, is exercised remotely by the distribution networks including switching inverters off. Shared controls between inverters provides a rich field for academic research where many control schemes are being modelled.
Technically most of the schemes are forms of droop control, basically borrowed from traditional synchronous generator operation but with a big difference. The big synchronous generators in power stations are operated by both control systems such as automatic generator control (AGC) and human intervention.
In microgrids, the schemes proposed either require a communication network connecting to all inverters and/or means to share droop controls. Things get very complicated, and—to be noted—without any practical examples. There are microgrids which avoid complexity by employing a master-slaves arrangement. For example, a large-capacity diesel generator or large-capacity battery-master inverter (voltage forming) and everyone else having grid-following inverters.
Related article: Barking up the wrong tree—an engineer’s perspective
Aggregation of solar owners into virtual power plants, or in fast frequency response schemes solves only a part of the conundrum. The tide of ever growing solar and batteries, unregulated as it is, has some of the aspects of a gigantic uninterruptible power supply—without being one. Distribution utilities cannot cope with this without drastic engineering changes to their networks. The required technologies would make Australia a leader. To achieve this a master engineering plan would have to be initiated and energy markets designed around it—not the other way, because that has proven not to work.
