By Phil Kreveld
An about-face is urgently needed in which system engineering takes absolute priority, and the development of markets is based on system engineering outcomes.
The latest wrinkle in electricity market design, as headed by Dr Tim Nelson, invites yet another appraisal of the physical state of the national electricity systems. Why the interest in the physical? Well, if it breaks, we’re all in trouble, even if we have ‘you-beaut’ markets!
Let’s forget about CO2 emission. That will do away with arguments from climate change sceptics as well as those of solar and wind enthusiasts bent on saving the planet. Whatever the reasoning was behind the present generation, transmission and distribution system, let’s use it as a basis for figuring out (1) how to ensure reliable supply of electricity as more renewable sources and batteries are inserted, and (2) at low cost to the national economy.
An ‘engineering first’ approach raises the question—why? We are accustomed to the economists and policy wonks having the first go. Yet these undoubtedly intelligent folk have gotten us into a miasma of claims and counterclaims regarding the ideal energy solution at each new turn of the ad hoc developments marking the transition path of the national electricity system. ‘Ad hoc’ or ‘as the situation demands’ interventions have gotten us into a situation where other than for renewable energy targets, we do not have a matching engineering plan that accommodates the technology challenges posed by the renewable targets.
Related article: A kind of knitted electricity system
The Australian Energy Market Operator’s Integrated System Plan, although an excellent document describing timelines for the addition of renewable energy sources and transmission lines, only hints at the engineering hurdles to clear. This article goes into some detail—and hence the title, because rather than getting trapped by ideology (wind ‘good’, solar ‘good’, gas ‘good/bad’, coal ‘bad’, nuclear ‘terrific/terrible’) the fact is that we can trot out solar and wind farms at speed, unlike pumped hydro and nuclear energy solutions but in order to make it all work there are engineering matters to be ironed out.
As to (2)—lowest cost, opens a can of worms, the first one out of the can being the ‘low cost to whom’ worm. The low-cost argument is artfully mixed with technology. Thus, there’s cheap technology and there’s expensive technology. The proponents from opposing sides of politics hold that the opposite corner has the expensive brand. The only rational way out is a protracted one, requiring research based on defined physical systems, straining objectivity, no matter how genuine.
As to (1)—the engineering aspects; there are enough angles to make it time-consuming but in the tasks of identifying and quantifying the engineering requirements, cost allocation becomes a realistic task. A regrettable matter is that electrical energy debates in the public square do not include engineering voices of those at AEMO, and in generation, or transmission, and distribution networks. Some readers may find the following stultifying, or simply unintelligible but this article is a genuine attempt to launch the very necessary engineering voice on the complexity of the renewable transition path. It will help focus on some baffling aspects such as the growth in rooftop solar and batteries, while concurrently planning more transmission and large-scale generation—and will hopefully invite reflection on electricity market designs that suit wind, solar and battery technologies.
State of play
The technical challenges in engineering terms are posed by the conjunction of synchronous and asynchronous energy sources, the decline in grid system strength as more asynchronous energy sources connect, and the control aspects of 100% asynchronous source electricity systems. Synchronous sources are rotating electromagnetic machines; asynchronous sources are essentially inverter based, whether for solar, wind or battery energy sources.
(1) Interest in keeping coal-fired power stations going is declining unless there is government intervention. This historical source of synchronous generation has provided system strength by virtue of its overload capacity of between 4 and 6 times of rated power capacity and although not mentioned in the ‘nuclear pitch’ already raises an interesting option: cost of nuclear energy-powered synchronous generation versus the cost of re-engineering coal plant to high Carnot-efficiency operation. With CO2 emission being a consideration, high efficiency, raised temperature operation, would reduce emissions from 1,000kg/MWh to perhaps half that value but not to zero. Both technologies provide synchronous generation that facilitates transmission grid strength and voltage phase angle and frequency stability. Grid strength as measured by short circuit current levels allows the use of existing circuit protection (reclosers and circuit breakers).
It also simplifies control aspects of the overall electricity system in that phase-locked loop inverter-based resources can be deployed to back up shortfalls in energy requirements and this relates to the limitations associated with IBR in contrast to synchronous machines. With sufficient synchronous generation in the grid the renewable solar sources and system IV wind generators can utilise PLL-grid following IBR, reducing control complexity and utilisation of existing protection devices. System III wind generators, which use AC excitation also operate better in strong grids thus making a good case for the retention of synchronous generation capacity. However, unlike battery sourced IBR, synchronous machines powered by steam, whether coal or nuclear, require a two-shift approach whereby generators can be idled while steam continues to be instantly available. This method is already practised and basically allows for synchronous generation to be available during absence of solar and wind energy, and during periods of high variability in wind/solar by way of spinning reserve mode. Note that market mechanisms would have to accommodate these modes of operation.
There is a degree of confusion about the functioning of PLL-based IBR and grid (voltage) forming IBR. Essentially the only difference is in their control aspects. For PLL-IBR, the d-q transformation of grid voltage at the point of connection, provides the phase angle of the current being delivered into the PoC. High grid strength is desirable—not for circuit protection, but for elimination of voltage and power oscillation that has been plaguing solar farms in Victoria’s north-west region (dubbed the Rhombus of Regret). The combination of synchronous generator capacity with PLL-IBR provides for the retention of conventional protection and voltage and power stability. The question remaining is the proportion of synchronous generation power that should be retained. There is a level of arbitrariness in this matter as the location of synchronous machines in relation to grid topology is highly relevant. However, a 30/70 SM/IBR ratio is a good opening assumption. Grid (voltage) forming IBR instead of a PLL, utilise current control to set voltage magnitude and angle. They can therefore mimic SM droop control (power-frequency, reactive power-voltage). In a zero-emission electricity future SM would be replaced by grid forming IBR. However, the questions of new protection and control schemes need to be resolved and that is a technical challenge as there are no practical examples to follow!
The role of rooftop solar generation is deliberately being ignored as a true agent of change. Instead, a conventional role is envisaged, i.e., as net consumer of electrical energy because that obviates the need for serious capital investment in distribution networks. However, purely because of existing market structures, the national policy is to pour billions into transmission lines in spite of the need for grid augmentation during periods of minimum power flow when suburban and regional energy duck curves are at their lowest points. In a recent podcast interview with Dr Tim Nelson, the matter of rooftop solar was raised. Suffice it to say, Nelson was not being caught in anything remotely controversial regarding the role of consumer energy resources. Instead there was a vague exposition of market flexibility schemes being considered.
Technically there are problems, which if resolved, could turn distribution networks into autonomous electrical energy centres. The technical issues are:
- the lack of uniformity in largely single-phase inverters,
- lack of a single communication standard for inverters,
- a non-uniform roll out of smart meters,
- d) an artificial division between meter owners and meter data owners,
- network insufficiencies in transformer behaviour under reverse power flow,
- the need for new voltage control methods,
- dealing with harmonics and phase imbalance and
- changes at terminal stations to deal with reverse power flow.
Distribution network owners have no financial benefit from autarky, so a larger degree of independence, though it might benefit CER owners and avoid much new transmission line construction is not on the agenda. In its stead emergency backstop measures are being considered to restrain CER energy outflow, whereby inverters are switched off in the interests of maintaining voltage stability.
The system comprising of long transmission lines, series compensators, shunt compensators including phase-shifting transformers, var and statcom compensators, synchronous condensers, and synchronous and asynchronous large-scale generation, large-scale storage and integrating this to provide voltage magnitude, voltage angle and frequency stability requires a ‘whole of system’ approach—precisely what we are lacking. This lack is leaving the door wide open to blame ‘renewables’ for grid problems when in fact we are not effectively understanding the system requirements of networks mainly comprising of IBR.
Related article: Solar facts, and the question of Australia’s net zero target
The peril we face is a that in having sought refuge in markets like frequency (and fast frequency) control ancillary services, FCAS, FFCAS, system restart ancillary services, SRAS, etc, to do what centralised control ought to do, for example CSIRO’s Control Room of the Future, developed in conjunction with the electric power research institute, EPRI, in the USA, we are piling ‘ad hoc’ complication on complication—and profit centre on profit centre in artificially created markets.
An about-face is urgently needed in which system engineering takes absolute priority, and the development of markets is based on system engineering outcomes. A good example is the voltage and power oscillation occurring with solar farms in conditions of low irradiance and in weak networks (lacking in short circuit current) and low X/R (inductance/resistance ratios) points of connection to the grid. Reactive power support in the form of synchronous condensers at the point of connection of a solar farm cost a motza, and may achieve no more than the reduction of voltage amplitude oscillation which was at perhaps a level of .02 pu. A whole of system approach may well find that we can live with that level of oscillation, and not trip power oscillation blocking relays.
A comprehensive control system must include distribution networks with visibility of medium voltage magnitude and voltage angle as well as detailed instantaneous power flow. Ideally this should be integrated with a granular short time interval power flow heat map of the low voltage portion. The latter would provide the basis for dynamic restraint of under frequency load shedding relays allowing net power outflow rooftop solar systems continuing to support low frequency events which would otherwise have resulted in the operation of the UFLS relay. The above scenario could apply as well to terminal station control where, ideally battery based energy storage and large rating IBR could be located to provide virtual energy independence to distribution networks.
In truth, we have a way to go!
