By Phil Kreveld
Janus, the Roman god of portals, doorways, war and peace—and of advantages hiding disadvantages—requires that we have another look at pushing green electrons around the place.
The green transition appears unstoppable but Janus points us to the disadvantage attached to all that ‘free’ energy from solar and wind farms. Pushing the stuff around to where it’s needed is costly.
Janus hasn’t met with Anna Collyer of the Australian Energy Market Commission, but he should. It will then become evident that the zero marginal cost of renewable source energy has high network costs associated with it—forcing new ways of charging for electricity.
Putting prices on wind and solar photons was always an artificiality. Janus forces the ‘other view’—capacity (to generate when required) and the cost of building networks that determine cost.
Related article: Avoiding a costly split due to irreconcilable differences
We are stuck in old technology and old ideas
The absolute necessity of moving electricity any distance has gone; cast a glance at Aussie rooftops. Why is it then, that those solar systems still need to be connected via distribution networks to high voltage transmission lines.
The obvious answer is that the phase-locked loop solar inverters require stable AC voltage supplied by distant generators via HV transmission lines. The question, however, is a disturbing one because the answer is ‘we’ve always supplied voltage (and power) that way—and even, if increasingly, no power is needed, but voltage is still essential.
Voltage—or system strength is de ‘mot de jour’. And EnergyConnect, joining South Australia to New South Wales, is not only a transmission line for South Australian energy heading east but also to furnish ‘system strength’ to ElectraNet. Were SA to island for 24-hours, the state would not be able to provide stable voltage to ‘keep the show on the road’ for that period.
It is obvious that the opportunity to rethink electricity transmission and distribution has not been taken advantage of. Janus would have us recognise that we have fallen prey to a completely inadequate economic model that prices kilowatt-hours from ‘freebee’ electrons. Not only that, but it’s forcing us to engineer expensive hardware around it—‘no transition without transmission’.
We are still making inverters kind of look like synchronous generators, and sending power from distant generators to distribution networks that are for many hours quite independent, energy-wise if not voltage-wise. Supplying mainly voltage in long transmission lines requires synchronous condensers, var compensators and synchronous generators operating in under-excited mode to suck up leading reactive power due to transmission line charging current. ‘Same-old, same-old’ boring recitation of facts but no solutions—therefore let’s do some peering into the future.
An inverter-based resource grid is a theoretical possibility. Testing a grid-forming IBR by connection to an infinite bus (or close to) is as far as we are going at present. However, it is not a realistic test for its operation in an IBR-populated grid, because system strength, i.e., voltage required to stay put, will have weakened considerably and neither small signal stability nor transient stability can therefore be adequately tested.
The Australian Energy Market Operator now has a new board member from Tesla and although adding to expertise, neither Tesla nor all the other inverter manufacturers who are part of the UNIFI inverter consortium appear to be close to agreeing on whether universal standards for grid forming (GFM) or individual technology developments should ‘set a flexible standard’ so as to allow for individual enterprise excellence.
AEMO might impose connection standards but meeting those in practice is another matter and it remains questionable whether those standards are adequate. Keeping technology improvements close to the chest of individual inverter suppliers is to be expected—and it is the reason why extensive flexible AC transmission systems (FACTS) remain an interesting but academic subject.
Standing in the way of a secure IBR-based grid are:
(a) Enforceable small signal and transient stability standards
(b) Enforceable reactive power support standards
(c) Protection relay standards geared to reduced fault current, sequence currents and phase jumps.
(d) Lack of a whole-of-grid control system.
These are major items, and in their absence, we return to the world of synchronous machines, The one thing we cannot afford to do is experiment with a vital piece of infrastructure—and that is AEMO’s dilemma. It is the one reason why the operator is counting on the installation of fifty and very likely more, synchronous condensers. Some will be provided by decommissioned power stations; the bulk will have to be purchased with minimum delivery times of three years and probably longer.
From frequency to no frequency markets
Returning to the two-faced nature of Australia’s national electricity system, it is obvious that the energy market must disappear in favour of a market for power (i.e., capacity) and network costs. A market mechanism for frequency control is anachronistic—invented in times when primary frequency control by means of turbine governors was disabled so as to allow scheduled energy dispatch. Governors have been reinstated but the fast and slow ancillary frequency markets remain—even though synchronous generation capacity is shutting down.
Were we to grasp the nettle, the frequency control ancillary markets would go because with more grid-forming IBR taking the place of synchronous capacity, the frequency band would widen. The main reason for this is that proportional power-frequency droop control, creating frequency offsets, would prevent grid forming IBR from fighting for frequency supremacy with each other which would happen if they were all equipped with proportional-integral control. A wider frequency spread would not be the end of the world as we know it but it would have to be figured in for grid operation stability criteria.
Someone will have to pay for reactive power
Voltage—to push green electrons around—is a serious matter. A shortage of reactive power sets the stage for voltage collapse. Providing reactive power is anathema to energy markets which only make money from supplying active power over time. It came free of charge with synchronous generators, but no longer—it is now supplied by synchronous condensers, presenting billions of dollars in grid augmentation investment.
In order to supply an energy consumption centre (load) via a transmission line, a generator must supply reactive power, needed by the line and load, each with separate requirements. The longer the transmission line the more reactive power support is necessary—and the flexibility necessary to change from inductive to charging current support. Inability to supply reactive power by a generator, collapses the voltage at the point of connection.
Time was when all concentration on reactive power requirements was met at load centres, i.e., distribution networks—it was called power factor correction. The replacement of direct, on-line motors by inverter-driven motors has improved power factor to near unity. However, our penchant for long transmission lines to connect renewable, remote energy zones is bringing back the need for reactive power—and to add to the problem, rapidly varying from inductive, when supplying large active power loads to capacitive reactive power when no or little power is being transmitted. In the first case, voltage drop must be restricted and in the second, voltage rise.
Whenever the media crow about how rooftop solar is almost meeting the nation’s energy requirement, AEMO wants to turn off those solar inverters so as to force up power demand. The rapid variation in transmission line power flow makes for expensive, as well as troublesome voltage control. Synchronous condenser excitation control, from over-excited for high power flow, and under-excited for low power flow, has to be ‘nimble’—and is likely to fail the test, thus needing back up from static var compensation, and static synchronous compensators. It all adds very significantly to the ‘grid augmentation’ bill.
Related article: Negotiating the two-way grid
From energy to capacity markets
The change to capacity markets is staring us in the face—and steadfast refusal to recognise this is causing investment hiccups through negative energy prices. Tim Nelson’s Energy System Entry Mechanism will not allay the nervousness for new wind, solar and battery investment unless we switch to capacity markets, i.e., rewards for being able to supply, when called on, active and reactive power.
AEMO’s integrated systems plan calls for 300GW of active power capacity by 2050, and thus we may well be looking at 350GVA of total capacity including reactive power. Someone has to pay for this! Allied to capacity markets is the need to pay for grid augmentation. The overall result is that the cost of supplying electricity on demand, whenever required, whatever the hour of the day, whatever the season, whatever the temperature, etc., etc. is not heading to zero, no matter the enthusiasm on display from energy ministers.
Furthermore, market mechanisms such as one proposed for inertia are when all is said and done, artificial creations of profit centres and premised on misunderstandings of the technology implications of the renewable transition. Janus is forcing the ‘other view’; one we are reluctant to mention in ‘polite company’—as in plenary sessions of energy conferences.
The mums and dads, the business owners, and everyone else in Australia are going to have to cough up simply for being connected to the grid whether one is creating one’s own energy or not. The connection fee will have to include the insurance premium for ‘over-capacity’ and it, rather than kilowatt -hours derived from sun and wind, will be the major charge.
The insurance premium also goes to meet the necessity for central control with 3 to 6 seconds (equivalent to up to 300 cycles) of latency to make decisions being replaced by 100 milliseconds, or less—if we go all the way with inverters. Machine learning will be required, with human intervention reserved for ‘one second and longer’ timescales.
No point looking to Roman gods—we need to remove the scales from our eyes and wake up to the technological realities—and therefore economic implications of the green transition.
