The ‘greening’ of the NEM and disappearance of rotational synchronous energy not only demands a future with more battery storage and other storage resources but also a comprehensive, big data, real-time and near real-time information and extensive network monitoring system.
The NEM is becoming a unique experiment
The National Electricity Market’s physical layer, comprising of widely dispersed synchronous generators, long, often close-to-congestion limit transmission lines and several hundred thousand substations, many of them subject to highly variable power flows due to millions of rooftop solar installations, is unique in the developed world.
Comparisons with Denmark, Spain, Germany, and other European countries, which are also subject to large renewable generation, don’t serve as they are part of a highly meshed, continental network so that on an aggregated basis, renewable power as a ratio of synchronous power is much smaller than for the NEM where participatory states are encouraging more renewables.
Marrying legacy electricity networks with electronic power converter technology using solar PV, batteries or wind energy as primary energy input, can only be achieved through the control capabilities offered by the use of big data, providing visibility of generator buses and HV transmission nodes down to fringe of network.
We do not have this as yet but urgently need to implement a transmission and distribution- inclusive, and comprehensive information system because without it we are heading towards network instability.
Replacing synchronous generation with electronic power converters necessitates a shift in information system paradigms.
It is easy to become comfortable with the capabilities of power converters to mimic forms of synchronous inertia behaviour and on that basis to conclude that, other than for despatchability concerns, the proportion of rotational inertia in networks is not of much importance.
However, we do not live in a theoretical world of computer simulations, and there are no practical examples of large grids where rotational synchronous power is less than 70 percent of total demand.
According to modelling studies of the National Renewable Energy Laboratory (USA) in respect of meeting the forecast demand of continental USA in 2050, 80 percent solar and wind power penetration resulting in 45 percent energy contribution, would require battery power capacity of 60 percent of total USA power demand. This implies major investments in the future but so would replacement of legacy technology reaching the end of its economic life.
Phase-locked loops (PLL) of electronic power converters operating under various control algorithms to provide inertial and voltage forming response are subject to modelled frequency and phase angle oscillations that have not as yet been experienced in the real world. The potential transition from a large synchronous base to a small one is, therefore, causing a great deal of concern, and essentially sets the scene for ‘flying blind’.
The absence of a second-by-second, extensive, synchronised network information base, militates a highly conservative approach
for retaining a ‘safe’ margin of rotational inertia.
What the brave, new world of renewables is forcing in the first place, is extensive deployment of synchrophasor measurements in network nodes (the inset below provides some basic information on synchrophasor technology).
They are a very sensitive, short-time response indicator of frequency shift and voltage collapse, and therefore essential in decision making such as providing battery support in a timely fashion and, when necessary, initiating selective islanding to preserve stability when dramatic load shifts occur.
Synchrophasors plus power flow and voltage measurement, form the essential data indication network stability criteria.
Twenty years ago, and longer, extensive phasor monitoring would have been considered a radical, unwarranted exercise in the face of relatively long network latency periods. In the world of renewables, scudding clouds, driving rain, wind fluctuation, etc are causing latency to drop to minutes and even shorter periods in distribution and in transmission networks.
A new challenge for AEMO
Synchrophasors, power, reactive power, voltage, harmonics, imbalance at substation level and in the HV lines data, being on a GPS-synchronised basis, should be the inputs to a big date system forming the backbone of AEMO’s control as NEM operator.
For control in the absence of real-time and near real-time information, the only other path is by way of prediction of demand though weather forecasts and previous day data furnished by distribution companies but that is ‘making do’ in the knowledge that a 24-hour control lag sits badly with network latencies of the order of minutes. Once the monitoring instrumentation and information system is in place, control strategies will evolve organically.
Gross investment in monitoring and data acquisition as the essential basis for network control is by comparison to contemplated investments in new HV links, increase of REZ link capacities, augmentation by way of synchronous condensers and SVCs, a close to insignificant fraction.
There is also no denying that investment in network assets without a highly granular monitoring and data system can lead to misapplied and even unnecessary funding.
The question is who should be the initiator and funder as generators, transmission and distribution companies will be party to the benefits conferred by the monitoring and big data system. The lead should be taken by the NEM operator, AEMO, who, in consultation with the AER, will allow the latter to regulate allowable investment by the network companies.
At the engineering level there will be necessary works, although not involving new technology. Edge computing on-board of monitoring instrumentation and more extensive deployment of IEC 61850 conforming IEDs as well as DREDs for the control of domestic solar inverters, will be necessary.
As to inverters, the laisse faire climate of, in effect minimum regulation, what has been allowed to be connected needs to be tightened, the reason being that power and reactive power, and islanding will have to be controlled by poles and wire companies with the overriding responsibility for network stability rather than various demand aggregators and VPPs.
The benefits for the paradigm sketched above will be tangible. Unforeseen grid events will be able to be dealt with while preserving network stability for the majority of the grid.
‘Keeping the lights on’ has been mouthed innumerable times by people who have little or no interest in the physical requirements that the statement entails but instead use it as an argument for maintaining the status quo.
The future, as envisaged here, will help a great deal in keeping the lights on.
Imagine a major fault event, causing a separation of an entire NEM portion. Islanding of many substations will occur and with that, through the use of monitoring and big data systems, intelligent control of voltage forming inverters within distribution networks will keep the lights on, and provide resynchronisation once grid restoration has taken place.