By Phil Kreveld
The renewables transition can be regarded in two ways: nothing to see here, or a big headache.
The first option assumes a 100% renewables grid close to a fait accompli; the second can’t see the finish line for all the hurdles in between. We will most likely end up falling between the hurdles, with the finish line receding.
An engineering problem that dares not speak its name
The Australian Energy Market Operator (AEMO) has a big headache. They don’t talk in public about the engineering problems of the transition from traditional to renewable generation. They are not at all sure that a 100% wind, solar, and battery grid is a reality. The safest way of managing the transition, they think, is to make sure synchronous machines never disappear from the grid. Therefore, as synchronous generation (SG) closes down, synchronous condensers (the same device as a generator but lacking a mechanical power source) are installed. The timelines for the transition thus depend on delivery schedules of syncons.
To be clear, a future grid with at least some 30% of grid-forming inverters (GFMI) to replace coal-fired generation and maybe some gas turbines is not fairyland provided transmission grids are equipped with syncons. It represents a massive investment. Adding this to the regulated asset base of grid operators assures the punters that as energy production costs go down, network costs go up, i.e., electricity tariffs will not go south any time soon if ever.
Related article: Renewables vs fossil fuels: Why can’t we walk and chew gum at the same time?
Replacing synchronous machines with synchronous machines
Syncons boost performance of GFMI if placed in tandem. Other syncons will be installed at various connection points (buses) in the transmission network, usually transmission line substations. Syncons can provide high circuit current (allowing reliable operation of circuit breakers when lines short circuit), voltage support and inertia. These roles fall under the heading of providing grid strength. All of this was the province of synchronous generators, now being pensioned off as coal-fired power stations vacate the field.
Transmission lines are not extension cords
As commonsense dictates, the longer a transmission line is the less power can be transmitted through it, all other pertinent factors being equal. Or, another way of putting it is that grid strength weakens as a transmission line lengthens. More accurately, the impedance (to power flow) increases with length. This is a challengeable statement but not if distances don’t exceed 200 or so kilometres. Much renewable generation connects to existing grids, often via significant impedance spur lines, connecting to transmission line corridors.
Although the point is often made that renewable energy sources cause weakened grids, it is not correct in that if an existing line because of high impedance is weak, it remains so irrespective what type of generation is connected to it. However, a weak line can have a deleterious effect on voltage stability—and in general, on generator stability.
Minding the volts and keeping the lights on
Voltage stability and generator stability are very important. If voltage stability cannot be maintained, the lights will go off, irrespective of available energy. Likewise, generator instability can bring a network down. The transition to renewable sources splits between a majority grid followers (GFLI) and a minority of GFMI. The higher the grid impedance, the more stability of GFLI is threatened. The reverse applies to GFMI.
Inverters only? A new ballgame
The grid followers are a passive lot. Provided they are connected to a steady voltage circuit they are a very reliable source of energy. In the best of all possible worlds, all renewables including batteries would be grid following with voltage supplied by synchronous generators. High-efficiency coal and gas fired generators would provide voltage-frequency, and would be supplying steady loads like Tomago, and be capable of working on a two-shift schedule.
Grid forming inverters have a major disadvantage when compared to SG. Maximum current and excitation voltage are limited compared to SG, rendering GFMI performance subject to presently unfamiliar grid topology. Take as an example a GFMI and SG sharing a common load bus. Both generators are to share in supplying active and reactive power. Active power is shared on the basis of the inverse of their power-frequency droop constants. Likewise reactive power is shared as the inverse of their reactive power-voltage droop constants.
We soon run into a problem if there is an increased demand for reactive power (in order to prevent voltage collapse). The SG simply increases excitation voltage, while maintaining active power. It has no problem with increased current due to additional reactive current being drawn. It sets the load bus voltage and the GFMI by virtue of its equivalent excitation voltage being limited by the battery-DC link voltage, drops its participation in reactive power provision. Furthermore, it has to work against the load bus voltage while being current-limited. Unless equipped with virtual impedance, the GFMI will be subject to ‘windup’ trying to maintain its active power contribution—and lose synchronism.
Related article: In defense of AEMO and its sorrows
Technology and topology in harmony
The above example illustrates the need for grid topology analysis as a result of incorporating GFMI. Where they are placed, and in combination with either other GFMI and SG will determine their power-reactive power performance, in addition to small signal and transient stability behaviour. If a syncon were connected to the load bus, both the SG and the GFMI would be relieved of having to supply reactive power. This also shows the desirability of network planning for inverter-based generation.
Although a gross simplification, grid weakness, in addition to making fault detection more difficult, also contributes to voltage and power instability. In the case of grid-following IBR, the higher the grid impedance to which they connect, the more likely sub-synchronous oscillations will occur. The reverse is the case for grid forming IBR (GFMI). However, with the bulk of the generation being grid-following, weak grids pose particular stability problems.
None of the foregoing indicates that we are on a hiding to nothing as regards the switch to solar, wind and batteries. However, we are badly served by purists who would add more hurdles to what is already a truly difficult trajectory. It is further aided and abetted by political skirmishing, causing puzzlement in investors.
