Batteries audition for the grid's safety net, the deep sea leaks its carbon, and where a regulator's power ends

The Australian Energy Market Operator, which runs the National Electricity Market across the country's eastern states, published its annual General Power System Risk Review on 31 July.[1] The document flags the loss of this fault current as one of its priority concerns for the coming years, alongside the strain that large new data-centre loads place on the network.[1][4] Its finding on the batteries is deliberately cautious. Grid-forming inverters have shown they can hold a stable voltage, it says, but they have not yet been confirmed to deliver, at scale, the fault current that protection equipment relies on.[4]

How the surge works01

System strength is the engineer's term for a grid's ability to hold its voltage steady and recover cleanly when something goes wrong, and it is gauged by the size of the fault current available at a given point.[3] A synchronous generator supplies that current as an unavoidable by-product of the way it works, because its heavy rotor stays magnetically locked to the grid and drives current into a fault whether or not anyone designed it to.[3] The relays that decide which breaker opens are calibrated to see that large surge, and if it is too small they can fail to trip or trip in the wrong place, letting a local fault spread.[3]

Inverters behave differently. The electronics that turn a battery's direct current into grid alternating current can pass only so much current before they must protect themselves, so they cannot match a spinning machine's overload.[3] Older grid-following inverters simply track the voltage the rest of the grid sets for them and add little strength of their own.[4] Newer grid-forming inverters actively impose a voltage waveform, which lets them steady frequency and voltage and even hold up a small islanded network on their own.[4][5] What remains unsettled is whether they can push out fault current of, in the operator's words, "sufficient magnitude, duration, and composition" to work the protection reliably.[2]

The trial02

To settle the question, the operator will gather real fault-current data from grid-forming batteries already on the system, one of several trials it is running under a mechanism in the National Electricity Rules for transitional services.[2] The stakes are more than academic. The market's battery pipeline is dominated by grid-forming designs, which feature in about three-quarters of a proposed fleet exceeding thirty gigawatts, even though their fault-current performance is not yet proven at scale.[4] Until it is, the fallback is the synchronous condenser, a large spinning machine that burns no fuel and exists only to supply strength and fault current.[4]

Those condensers are proving expensive and slow to build. In New South Wales the network company Transgrid has seen the cost of its first tranche run to about 1.1 billion Australian dollars, well over early estimates, and has proposed substituting 900 megawatts of grid-forming batteries for two of the condensers it had planned, if their capability can be shown.[4] The retirement calendar sets the pace. The Eraring plant in New South Wales, the country's largest coal station, is due to close by the end of the decade, and Victoria's Yallourn station is scheduled to shut sooner.[4]

A limit every clean grid meets03

The problem is not peculiar to Australia. Fault current and system strength fall on any grid where inverter-based wind, solar and batteries displace spinning generators, so the same challenge is surfacing on other systems moving quickly off coal and gas.[3] Australia is an early and severe test case because its coal fleet is leaving fast.[4] "If it can be proven that these batteries that are grid forming can support protection quality fault current," AEMO's Nicola Falcon has said of the work, "this is a really big game changer."[6] The trial's job is to replace that "if" with recorded fault-current data.

The work, published in the journal Science Advances by a team led by Peter Stief and Ronnie Glud at the University of Southern Denmark with colleagues in Germany and elsewhere, tackled a question that is hard to study in the open sea.[1] Marine snow sinks fast, and as it falls the water pressure climbs by roughly one atmosphere every ten metres, reaching hundreds of times surface pressure in the deep ocean.[1] To recreate that descent, the researchers grew artificial marine snow from clumped diatoms, the microscopic algae whose sticky aggregates make up much of the real thing, and incubated it in rotating tanks that held the particles in suspension while raising the pressure to match depths of two to six kilometres.[1][2]

The squeeze01

Pressure alone proved enough to force a large share of the particles' contents out as dissolved organic matter, the loose molecules that leach from a particle rather than staying bound in it.[1] At pressures matching the deep ocean, the aggregates shed dissolved carbon equal to about half of what they had started with.[1] They gave up an even larger fraction of their nitrogen, on the order of six in ten, so what leaked out was rich in nitrogen relative to carbon.[3] Analysis of the freed material showed it was made largely of proteins and other easily digested compounds rather than tough, weathered remnants.[1]

That composition matters because it decides who eats it. When the researchers offered the leaked matter to a community of ocean-water microbes, the bacteria consumed it quickly, multiplying about thirtyfold within two days and leaving behind only the harder-to-digest leftovers.[3] Stief likened the effect of the pressure to a kitchen tool. “The pressure acts almost like a giant juicer,” he said. “It squeezes dissolved organic compounds out of the particles, and microbes can use them immediately.”[3]

A leakier pump02

The finding revises a standard picture of what scientists call the biological carbon pump, the process by which surface life packages carbon into sinking particles and ships part of it to the deep sea, where it can stay out of the atmosphere for centuries or longer.[1] Researchers had largely attributed the carbon lost on the way down to microbes and animals feeding on the particles.[4] The pressure effect is different in kind, a physical squeezing rather than a biological one, and it means a particle can be stripped of much of its carbon without anything having eaten it.[1]

Carbon that leaves a particle as dissolved matter and is respired by microbes in mid-water never reaches the seafloor, so less is buried than a model tracking only intact particles would predict.[1] The authors are careful about how far the result reaches. They write that the process “likely weakens” the pump, a direction rather than a settled quantity, and they put no number on how much less carbon the whole ocean buries as a result.[1] Turning the laboratory measurement into that global figure, and into the ocean models used to project the carbon cycle, is the work that remains.[4]

The case reached back to a cable-television fight in Tamil Nadu.[2] In 2008 a multi-system operator, the middleman that carries channels to neighbourhood cable networks, cut off its signal to four local operators during a quarrel over subscription dues, without giving the notice the rules require before a disconnection.[2] The Telecom Regulatory Authority of India, known as TRAI, found a likely breach, directed the operator to switch the signal back on, and issued a show-cause notice when it did not.[2]

The operator took the matter to the Telecom Disputes Settlement and Appellate Tribunal, the specialised body Parliament created in 2000 to hear disputes between telecom players.[1] The tribunal sided with the operator, reasoning that in ordering the signal restored amid a commercial quarrel, TRAI had effectively decided a dispute between two service providers, which is the tribunal's exclusive job.[2] TRAI appealed to the Supreme Court.[1]

The line the court drew01

On 24 July a bench of Justices S.V.N. Bhatti and N.V. Anjaria set the tribunal's decision aside.[1] Writing for the court, Justice Bhatti held that a direction to obey a regulation is an exercise of TRAI's regulatory function, not a determination of a private dispute.[1] The order had resolved nothing between the operators, the judgment reasoned, because it awarded no damages, fixed no arrears and adjusted no accounts between them.[1] A show-cause notice decides even less, the court added, since it only puts its recipient on notice and leaves any actual penalty to be decided elsewhere.[1]

Where that penalty is decided is the hinge of the reasoning. Under the statute TRAI cannot punish a breach of its own directions itself; it must complain to a criminal court, and only a magistrate can impose the penalty.[1] Because the regulator's order sits upstream of that process and settles no one's rights, the court treated it as regulation rather than adjudication.[2] The dividing test, the judgment said, is the “true substance” of the grievance: where that substance is a dispute between two providers, the tribunal has jurisdiction; where it is the breach of a regulation, the regulator may act.[1]

Why it matters02

The distinction between regulating and adjudicating is old, but its edge had blurred.[1] Regulators everywhere issue compliance orders against firms that are also caught up in commercial disputes, and a rule that any such order becomes an off-limits adjudication whenever a dispute lurks nearby would hollow out the power to enforce.[3] By tying the label to what an order actually does, rather than to the mere presence of a dispute, the court preserved a workable boundary.[1]

For TRAI the practical effect is to confirm that it can order companies to follow its rules, in telecommunications, broadcasting and cable alike, without being pushed aside to the tribunal every time a contract sits in the background.[4] The outer limit stays in place. The regulator still cannot award damages or settle the commercial claim itself, and the sharpest sanction for defiance runs only through a criminal court on the regulator's own complaint.[1]