Green hydrogen: mitigating electrical instability in off-grid networks

Using off-grid renewable energy to produce hydrogen

Electrical design challenges with green hydrogen production on off-grid networks and solutions to ensure a stable and secure power supply.

Electrical design for off-grid power systems

As green hydrogen projects grow into the GW scale, new technical challenges arise.

Congestion on the power grid makes connection capacity scarce, and the high costs of connecting remote regions are encouraging some projects to consider an off-grid design. In these systems, renewable power is generated by wind turbines or solar photovoltaic (PV) panels and then converted directly into hydrogen or ammonia for export without transiting a utility grid.

This allows hydrogen to be produced in remote areas with some of the best and cheapest renewable energy resources, without a requirement for grid access.

However, maintaining the stability of these off-grid networks can be technically challenging and poses some unique design challenges.

Challenges

This whitepaper explores the challenges of using off-grid renewable energy for green hydrogen generation, in particular around three key questions:

• What are the technical challenges with off-grid networks?
• What factors should be taken into account during design to mitigate these challenges?
• How can energy storage solutions contribute?

Balancing renewable energy supply with production demand

Learning from microgrids

Off-grid power systems are a somewhat mature and well‑understood technology. To date, almost all real‑world implementations have been microgrids. As the name suggests, these microgrids typically have a relatively small level of demand and generation. Their definition in terms of size varies, but in the US, the Consortium for Electric Reliability Technology Solutions (CERTS) defines them as less than 500 kW[1].

Microgrids have traditionally relied on diesel generators for their inherent ability to form and maintain a stable network as the primary source of power supply. However, there are some examples of microgrids that have high penetration of renewable energy generation, such as Borrego Springs in California [2] or The Isle of Muck in The UK [3].

Scaling this isolated technology beyond the microgrid level and into the GW scale makes the design significantly more complex. At the GW scale, design features of both microgrid and utility grid systems must be considered.

Voltage stability classification in microgrids

Balancing power in large, islanded grids

Utility grid systems rely on sophisticated commercial and technical mechanisms developed in recent decades to ensure that supply and demand are always balanced. Unlike these extensive networks, a large-scale islanded system would depend entirely on local generation to meet demand levels.

Crucially, the system must consider the inherent unpredictability of renewable sources like wind and solar PV, which typically have capacity factors below 50%. Electrolyser systems on the demand side can be optimised to have an inherent level of flexibility, which enables them to match the output from renewable generation and balance the system.

Fault level issues

The problem with inverter-based generators

In large power networks, electrical faults are inevitable. Faults can be caused by several things, including equipment malfunction, adverse weather or human error. Synchronous generators provide their respective grids with large, short-circuit currents during fault conditions. This arises from the inertia of large rotating masses. During a fault, some of the stored kinetic energy in the rotor is extracted and supplied to the grid, resulting in a larger short circuit ratio and, thus a grid that is more resistant to change. Inverter-based generators, such as modern wind and solar PV, inject much lower currents during a fault.

As a result, islanded grids with inverter-based generators have much lower short-circuit power. The protection systems in an off-grid environment need to work with low fault conditions and new solutions may be required to enable safe operation. Low fault currents make it harder to detect faults at all and generally lead to poor voltage regulation.

There are some advantages to low fault currents. For example, if the protection systems failed and a fault was experienced by electrical equipment, lower fault currents could mean less severe damage to equipment.

Overcoming short circuit-level issues

Modern, large wind turbine generators (WTGs) mostly deploy fully rated converter topologies, meaning there is no direct mechanical coupling. The converter can still inject current into the network under fault conditions, and this simulates the response of the large rotating masses in synchronous generators.

However, as it requires an active measurement and response from the control system of the WTG, it is not as fast as the response from a true synchronous machine like a synchronous condenser, which further reinforces the necessity of synchronous condensers in off-grid systems.

Frequency control

Managing frequency fluctuations

In an islanded electrical system, where intermittent renewable energy sources are the only generators of electrical power, a well-designed balancing mechanism would be required to address any fluctuations in frequency. The frequency of an electrical grid is related to the balance of active power by the swing equation. Frequency stability depends on matching demand with generation. If the demand exceeds generation, there will be a reduction in the active power, the frequency will drop, and vice versa.

Controlling grid frequency

Traditionally, utility grids relied on synchronous generation sources such as gas turbines. These generators operate at the grid frequency, and since their fuel source is controllable, they can quickly increase or decrease their outputs to rectify any change in grid frequency. As wind and solar are intermittent energy sources, their output will always have a degree of variability, leading to reduced frequency stability.

Importance of inertia

Synchronous machines directly connected to the network, such as diesel generators, induction motors or synchronous condensers, have a spinning driveshaft with rotating inertia. The driveshaft’s rotating frequency is directly connected to the system frequency according to a ratio determined by gearing and the pole pairs present in the machine.

The rotating driveshafts act as a temporary store of energy for the system; if there is too much power coming into the system, the imbalance will be used to accelerate the driveshafts, increasing the frequency. If there is insufficient power generation, the imbalance between demand and supply will be made up with power taken from the rotating driveshafts and the frequency will fall.

The rate of change of frequency (RoCoF) of an electrical system is directly linked to the amount of inertia the generators have. The higher the inertia is, the slower the frequency will fall when generation reduces.

Missing mechanical inertia

A problem for a green off-grid system is that renewable sources utilise inverters for power generation and rely on advanced control systems to synchronise with the grid’s frequency. Inverter-based generation does not contribute to system inertia, which is especially problematic in networks with high solar PV generation.

A utility grid consisting of large synchronous generators can rely on its inertia to slow the rate of the frequency drop. Older WTG models do have a direct mechanical connection to the grid and, similarly to synchronous generators, use gear ratios to match the grid frequency. These topologies would contribute mechanical inertia to the system. However, this is not the case in more modern topologies where fully rated converters are used in generation.

Lower fault levels and lack of inertia will both impact the stability in GW-scale off-grid networks.

Voltage control

Maintaining stable voltage levels

Utility grids worldwide use typical nominal voltage levels for the generation, distribution and transmission of electricity. Maintaining nominal voltage levels ensures efficient grids and reliable operation of equipment such as electrolysers.

Electrical transmission or distribution lines always have losses, but choosing and maintaining a suitable voltage for the application minimises this inefficiency. A well-designed islanded grid would have the advantage of building substations and electrolysers close to the generators, reducing electrical losses associated with transmission distance. Additionally, spreading the electrolysers across the network would mitigate the severity of outage risks. If the power supply to one electrolyser was lost, production could still continue.

Balancing reactive power

Meeting the demand for active power is insufficient on its own; an electrical network must also balance the reactive power. This means installing reactive compensation to compensate for reactive power sources naturally present in the system and operating some generators in voltage control mode where they regulate their reactive power output according to voltage.

Inverter-based generators and under-voltage protection

Inverter-based generators have protection systems that are designed to disconnect them in under-voltage conditions. This is to prevent damage to the equipment. However, this presents another challenge in off-grid systems. These systems have a fault-ride-through time that determines how long they will stay connected in the event of voltage deviations.

As islanded systems cannot rely on the wider network to rectify under-voltages quickly, this could cause generators to disconnect if the voltage does not return to nominal within a reasonable time. This protection feature could be made less conservative for an off-grid system, but this would come at the cost of significantly increasing the risk of damage to power electronic equipment.

Mitigations and solutions

The challenges outlined in this paper are not easy to solve in practice. Any successful network of this kind would have to be purpose-built with generation, distribution, transmission and all loads designed with consideration for every aspect of the project. Advanced control and protection methodologies would also have to be designed for such a system.

There are, however, several well-established and emerging technologies that would mitigate some of the presented issues.

Utility grids are designed to ensure security and quality of supply. In a system where generation and demand are being developed in tandem, the design would be quite different. For example, load shedding, where demand loads have their supply cut off to balance the grid, is treated as a last resort in most utility grids. An off-grid renewable energy system can be more resilient to supply cuts.

Inverters

Inverter technology is pivotal in renewable energy systems. In solar PV, it converts generated DC power into AC, while most modern turbine topologies change their generated AC into DC using a rectifier, then use an inverter to change it back to AC to export.

Inverter systems within generators can broadly be categorised as either grid-following or grid-forming. The majority of installed inverters, including on wind and solar PV generation sites, are grid-following inverters. These synchronise to the frequency and voltage levels of the grid and will disconnect generators in the event of frequency and voltage events, as described in the previous paragraphs.

An alternative type is grid-forming, which works to impose its own frequency and voltage within a network. Once established, other grid-following inverters can synchronise to this frequency. Rather than having a fixed power setpoint, grid-forming converters vary their power output to maintain the desired frequency and voltage.

Inverters can also provide synthetic inertia by emulating the behaviour of traditional synchronous generators. This capability is critical in managing sudden changes in load or generation by adjusting power output in real time. Synthetic inertia is crucial in environments with high renewable penetration, where physical inertia is inherently low due to the predominance of inverter-based resources.

Synthetic inertia requires a measurement of the system frequency and a control system response to the changing frequency that emulates the response a machine with actual inertia would provide. This requires drawing power from a source which could be a battery or rotating wind turbine blades.

Because a control system response is required, synthetic inertia is a bit slower than natural inertia, and some natural ‘true’ inertia is usually required, which can be supplemented by synthetic inertia within 50ms to 1s.

Energy storage solutions

When renewable energy is insufficient, traditional utility networks rely heavily on the burning of fossil fuels to bring voltage and frequency back to nominal levels. Smaller microgrids often use diesel generators as well. However, green hydrogen should be produced without any reliance on carbon-emitting generation methods.

An off-grid hydrogen producer would have to match the hydrogen generation to the available renewable energy generation.

Temporary issues arise when the maximum ramp rates of generation and demand are not matched. Energy storage presents a solution. It can act as a buffer to rectify frequency drops by releasing active power quickly. More advanced systems can also supply reactive power to help maintain voltage.

BESSs RoCoF protection

Battery energy storage solutions (BESSs) can provide synthetic inertia to an islanded grid. Unlike utility grids, inverter-based generation does not have inherent mechanical inertia from the synchronous generators. This synthetic inertia helps mitigate the issue of high RoCoF events.

Energy storage solutions also help to mitigate the more fundamental issues with renewables; these systems can recharge when generation exceeds demand and discharge when demand is high. This means, that for a short time, hydrogen production could continue in adverse weather conditions.

Static VAR compensators and synchronous condensers

The static VAR compensator, specifically a static synchronous compensator (STATCOM) can control voltage changes. These devices can rapidly change how much reactive power they supply or consume and are commonly used to support a grid under fault conditions. Their flexibility to provide either inductive or capacitive volt‑ampere reactive (VAR) makes them especially useful in voltage and power factor control.

In addition, to mitigate against challenges associated with low-fault current levels, short-circuit fault current can be boosted by a synchronous condenser. This consists of a motor connected to a generator, sometimes with a flywheel to increase rotating inertia, and it is one of the most common pieces of equipment to improve short circuit levels.

Conclusion

Islanded grids for green hydrogen production are feasible with careful electrical design.

Producing green hydrogen on islanded grids is technically challenging and requires a well-engineered electrical system for stable and economic operation. Paramount to the successful development of such a system is a well‑planned design and selection of components on both the demand and generation side.

This paper demonstrates that, by using grid-forming inverters and energy storage solutions, powering islanded grids with 100% renewable energy sources can be feasible and commercially viable.

Grid-forming inverters allow renewable generators to simulate some of the inherent advantages of synchronous generators that utility grids rely on. Advanced control methodologies and other devices can help manage frequency, voltage regulation and fault current level issues. In addition, energy storage can act as a buffer for any drops in power output caused by weather.

Although the cost of building such a network would be high, a well-designed GW-scale islanded grid in the correct location could be economically feasible in context of the expected rise in global demand for green hydrogen.