Rebirthing coal power stations into synchronous condensers

As the power system transitions to one dominated by Inverter-Based Resources (solar cells, batteries and wind turbines), the Essential System Services (ESSs) previously supplied by conventional technologies must be replaced. A key activity is the rollout of Synchronous Condensers (SCs). SCs are essentially a generator such as would be found in a conventional power station, spinning with the grid’s frequency, but without being attached to a power station.

SCs have played a niche role supporting local grid pockets for decades. However, power systems around the world are now scrambling for them in great numbers. Unsurprisingly there is a major backlog of orders, and they are far from cheap.

There is another solution. Rather than ordering new equipment, existing generators in thermal power stations could be rebirthed into large SCs. In theory, this makes a great deal of sense. The generators are fundamentally the same type of machine as a SC, and the grid connection, switchgear and transformers are already in place.

However, when these power stations were built decades ago, their designers never allowed for this new role, and there are large technical and practical challenges in retrofitting them. And in some ways they could never perform as well as a purpose-built SC.

ARENA has supported and published a report into exactly these challenges and limitations, prepared by industry veteran, DigSILENT’s Tim George. This report is very timely because the queue for new SCs requires us to commit to either them, or plant conversions, long before power stations close. Either way, we could easily regret our decision.

Despite the technical topic, the report is written in a very accessible way.  We touch on some points below and hope it gives you cause to read DigSILENT’s words directly.

What is a synchronous condenser?

Conventional power stations use synchronous machines to convert mechanical energy to electrical energy. Some large motors, such as those that crush mine rock, also use synchronous machines to do exactly the opposite.

Figure 1: Cut-away of a synchronous machine showing a spinning rotor inside a fixed stator

Source: Technical Engineering School (online)

Synchronous machines are coupled to the power system frequency, meaning that the magnets in the spinning rotor pass the stator’s conductors fifty times a second. These magnets cannot go faster or slower than the system’s frequency.

If a power station turbine pushes harder into its generator, it will not speed up, instead the generator simply sends out more megawatts. Similarly, if you put more load on a synchronous motor, it will not slow down, instead it draws more electricity from the grid.

Synchronous machines, both in power stations and customers, do other things that help the grid:

• They provide inertia. As the spinning rotor is heavy, it resists changes in speed, and as it is locked to the frequency, it resists changes in that too. If the grid frequency starts to fall, which is a symptom of not enough energy going into the grid, for a few seconds the machine will pump out more energy, slowing the decline. Vice-versa for increasing frequency.
• They provide voltage control. By adjusting the rotor’s electromagnet, the stator will create or consume reactive power, a description of electrical current whose waveform is out of phase with the grid’s voltage. Just like energy, the grid’s reactive power must be kept in balance, and any imbalance results in the grid’s voltage going up or down. The rotor’s electromagnet is tuned such that the synchronous machine adjusts reactive power to drive the grid voltage back on target.
• They provide fault current. If the grid suffers a short-circuit somewhere, for example caused by a lightning strike, the machine will, for a brief period, feed in a very large amount of reactive power. This lessens the resulting voltage disturbance, so that the grid can safely disconnect the short-circuit and recover. In the National Electricity Market (NEM), this characteristic is described as system strength.

If you disconnect a generator from its turbine, or disconnect a synchronous motor from its load, it will continue to spin at the grid frequency. It will continue to supply all the beneficial ESSs above while drawing a small amount of energy from the grid to supply its frictional losses. This is essentially what a SC is.

As such, a SC is a synchronous machine with a grid connection but is not attached to a turbine nor the rest of a power station. It does need a way to start up, i.e. spin the rotor up to grid frequency so it can synchronise. This can be done with a “pony motor” that disengages after synchronisation. Alternatively, it can be started electrically via a frequency converter which feeds in a very slow and progressively accelerating grid frequency into the SC’s stator, accelerating its rotor until it matches the system frequency.

In Australia, SCs are usually purpose built, owned and operated by the monopoly network, funded by customers as part of network charges.

Figure 2: Components of a purpose-built synchronous condenser

Source: DigSILENT report

SCs were historically built for voltage control in local network areas electrically remote from large generators, typically in the distribution network. Because, unlike a generator, they are not connected to the heavy weight of a turbine, these SC don’t naturally provide much inertia and were not installed for that purpose. In the future power system inertia will be in short supply, so SCs are being built with heavy flywheels to increase their weight.  

The South Australian transmission company, ElectraNet, recently installed four large, heavy SCs into their transmission network, primarily to maintain system strength. The full project took over three years and cost $190m. Looking forward, DigSILENT anticipates lead-times of at least 30 months for delivery and installation, which may lengthen in the next few years as the backorder list grows. DigSILENT estimates approximately $35-$40million for each 125MVA SC. Larger SCs are available, but DigSILENT has prepared their work based on this standard size that the Australian Energy Market Operator (AEMO) uses in its documents.

Figure 3: Installation of ElectraNet Robertstown Synchronous Condenser

Source ABC

Under the NEM’s previous “do no harm” arrangements, renewable generators were sometimes required to install small on-site SCs before they were allowed to connect. Under the new System Strength framework, connecting renewables are sharing contributions to larger, discrete SCs.

Figure 4: Musselroe Windfarm on site Synchronous Condenser

Source ABB

An insatiable demand for system strength

AEMO, as independent Victorian Transmission Planner, have just announced a remarkably large and near-term requirement for system strength, primarily in the Latrobe Valley. 

Table 1: Equivalent network solution requirement to meet AEMO’s system strength standard

Source: AEMO Victorian System Strength Requirement Project Specification Consultation Report

AEMO notes:

“Due to timing constraints, AEMO Victorian Planner does not consider network solutions (conventional synchronous condensers) alone to be a credible option to meet the system strength standard, at least by December 2025.”

And

“Due to challenging supply conditions, AEMO Victorian Planning has estimated a three-year lead time for delivery of synchronous condensers, from an order being placed to completion of commissioning.”

Meanwhile the Queensland Energy and Jobs Plan has also identified a great need for SC capacity as the role of conventional plant declines.

Figure 5: Queensland proposal to convert coal plants to synchronous condensers

Source: Queensland supergrid infrastructure blueprint 2022

Power stations as synchronous condensers

It is straightforward and common for hydro power stations to operate as SCs. Having started a generator from its water turbine, the inlet valve can be shut off and the turbine allowed to spin in air with minor frictional losses. In that mode the generator can be run indefinitely as a SC, providing ESSs to the grid. This capability already exists in many National Electricity Market (NEM) hydro power stations.

However, with the transition the ESS requirements are much greater and widespread than hydro plant could support.

So, given coal-fired power stations are closing with excellent grid connections and synchronous machine generators, can they not continue to have a useful new SC life?

Unfortunately, it is nowhere near as easy for fossil-fuel plant to convert and operate as a SC:

• Firstly, a steam or gas turbine can’t spin in air at low losses like a water turbine. The turbine must be disconnected from generator or have all its blades removed.
• Secondly, if the power station is no longer operating, it needs a new way to start the synchronous machine with either a pony motor or frequency converter.
• Thirdly it needs new control systems to manage the machine as a SC as well as other ancillary matters, such as increased noise.
• Fourthly, as much of a generator’s inertia comes from its turbine blades, if the SC is to provide equivalent inertia, it will require a flywheel.
• Fifthly, the ongoing staffing and maintenance requirement of an ex-power station site and generator would be higher than an automated, purpose-built SC.

Any retrofit is also highly site dependent. For example, is there space for a new thrust bearing to enable disconnection from the turbine, and can the foundations safely carry the weight of a flywheel?

DigSILENT estimate that converting a coal plant to SC would cost 60-100% the cost of equivalently sized purpose-built SCs and take 18-24 months.

Unlike steam plant[1], it is realistic to retrofit gas turbine plant, such that it could, like a hydro plant, operate both as a generator and SC. It requires installing a clutch and new thrust bearing to disconnect the turbine from the generator. Conversion costs would be about 60 per cen of the equivalent SC and take 6-8 months during which the generator is unavailable. This could well be a promising way to meet part of the Victorian requirement.

Purpose-built SCs have a life of around 60 years, and DigSILENT suggest that the ongoing operational advantages (like lower staffing) would put them at an advantage over generator conversion when considered over the long-term. But another unknown enters the discussion. Developments in grid-following inverters, seem likely, in future decades, to render SCs obsolete. Hence DigSILENT suggests we should discount these long-term benefits of purpose-built SCs.

In the short-term DigSILENT recommends feasibility studies be taken into all prospectively suitable generators, which would each cost $350,000-$500,000 and take 6-12 months.

Market design issues

Like so many electricity market issues, converting of generators into SCs falls into the limbo world that lies between competitive markets and monopoly grids. The large ElectraNet SCs were built as part of the regulated monopoly asset base, with costs recovered from captive customers. However, the convertible generators are competitive assets.

As noted above, some small SCs have already been built by generators to facilitate their connections. In their Post 2025 review, the Energy Security Board (ESB) recommended that markets be developed to purchase ESSs from competitive assets, indeed a competitive inertia market is being introduced into the Western Australian Wholesale Electricity Market (WEM). If so, the competitive sector could entrepreneurially invest in the conversion in response to signals provided by these markets.

However, a recent Australian Energy Market Commission (AEMC) announcement anticipates a different direction than that envisaged by the ESB. Instead, the AEMC now considers that ESSs should be procured under long-term contracts between transmission companies and generators. This is also how DigSILENT anticipates a conversion would occur. However, there are challenges in such an approach:

• Because the two approaches have such different physical characteristics and risks, could a regulator be convinced that a for-profit transmission company impartially compares contracted SC services to its own assets? For example, the Victorian system strength plan, developed by its independent transmission planner, AEMO, seems more prepared to procure beyond conventional network assets.
• The losses required to run a generator as a SC, about 1-2 per cent of the generator’s capacity, are paid by the generator. Similarly, the losses of small connection SCs are paid by renewable generators. However, the losses in network owned SCs are socialised across the market as transmission losses, distorting the comparison.

Conclusions

Despite the practical challenges they have identified, DigSILENT remain upbeat in the concept:

“The option to repurposing existing generators as SCs presents a credible and attractive opportunity to meet AEMO's security requirements because:

• It is technically feasible and, in most cases cheaper, to convert existing generators for use SCs,
• The scale is such that one big existing generator (say 750 MVA) could obviate the need for five or more smaller SCs (say 125 MVA), and
• Conversion times are expected to be quicker than implementation of new SCs.

These conclusions make a convincing case to pursue the re-purposing possibilities, recognising that there is a significant variation across sites.”

There are complexities and risks, but in the current climate, there remain significant upfront cost and time savings compared to the alternative of a purpose-built SC. Like the man going to Dublin, if you want an easy way to get a SC, you should probably not start with a coal-fired power station. But from where we are now, this may be the way to Dublin.

 

[1] All coal-fired generators and some older gas-fired generators are steam plants. All gas turbines are natural gas or liquid fuelled. There is however no expectations of widespread gas turbine closures in the near future.