Contingency limited modular nuclear deployment in island power systems with an illustrative Dominican assessment

Island power systems face a unique and unforgiving challenge: they are small, isolated networks where a single large failure can plunge the entire grid into darkness. Unlike massive continental grids that can absorb the sudden loss of a power plant by drawing on distant reserves, an island system must balance itself instantly. If a generator trips offline, the remaining machines must immediately pick up the slack, or the frequency of the electricity will drop so fast that protective relays shut everything down to prevent a total collapse. For decades, engineers have debated how to fit new nuclear power plants into these fragile systems. The promise of small modular reactors is that they are smaller and safer, but a critical question remains: does breaking a large plant into smaller pieces actually make the grid safer, or does it just create more complex problems? The answer depends on whether those smaller pieces can truly fail independently of one another.

A team of researchers at the Instituto Tecnológico de Santo Domingo set out to answer this question with a rigorous, computer-based experiment. They did not build a physical reactor or run a real power grid; instead, they constructed a detailed mathematical model of an island system, using the Dominican Republic as a realistic scale for their calculations. Their goal was to determine exactly how much money a power company could afford to spend on making nuclear modules electrically independent, based solely on the savings gained from needing less emergency backup power. They wanted to know if the cost of separating the modules was worth the risk reduction, or if the savings would simply vanish once the modules became too small to matter.

The researchers began by defining a specific type of danger: a "common-cause" failure. In a typical power plant, even if the reactors are separate, they might share the same turbine, the same electrical switchyard, or the same control room. If a fire or a mechanical fault hits that shared equipment, it could knock out multiple reactors at once, creating a massive power loss that the grid cannot handle. The team asked a simple but profound question: if you split a 400-megawatt plant into four 100-megawatt modules, does the grid actually become safer? Their simulations showed that it only becomes safer if those four modules are truly independent. If two of them are still tied to the same switchyard and trip together, the grid sees a 200-megawatt loss, which is no better than having two 200-megawatt blocks. The safety benefit only appears when the modules are so well-separated that a single event cannot take down more than one of them.

Using their model, the team tested five different ways to arrange a 400-megawatt nuclear station. They compared a single giant block against two 200-megawatt blocks, and then against four 100-megawatt blocks with varying levels of independence. They simulated the moment a block failed, calculating how fast the grid's frequency would drop and how much emergency power would be needed to stop the collapse. They found that if the modules are truly independent, splitting the plant down to 100-megawatt units eliminates the need for an extra 54.28 megawatts of emergency battery power compared to a 200-megawatt setup. However, this saving only holds up to a point. Once the modules are small enough that their failure is smaller than the largest existing problem the grid already faces, making them even smaller provides no further benefit. The researchers called this the "saturation point," where the grid's needs are already met, and further subdivision is just an expensive exercise with no return.

The most striking part of their work was attaching a price tag to this safety. They calculated how much extra money a utility could spend on building those independent 100-megawatt modules, knowing that the savings in emergency batteries would eventually pay it back. In their central scenario, the math suggested that a power company could afford to spend an additional 61.08 US dollars per kilowatt of capacity to achieve that level of separation. This is a very specific, narrow financial limit. It means that if the cost of making the modules independent is higher than that amount, the investment does not make sense based on frequency safety alone. If the cost is lower, it might be worth it. But the researchers were careful to note that this is not a universal rule for all nuclear plants; it is a precise boundary derived from their specific assumptions about battery costs, reactor lifespans, and grid behavior.

To ensure their simple math matched reality, the team ran a second, more complex simulation that included the messy, real-world delays of mechanical systems and the lag in how quickly batteries can respond. They compared the results of their simple model against this more detailed one. The two models agreed closely on the central case, but they disagreed in 85 out of 3,645 different scenarios. In those disagreements, the simple model sometimes said a design was safe when the complex model said it was not, and vice versa. This revealed a crucial limitation: while the simple math is useful for a quick check, it can be fooled by specific combinations of slow mechanical responses and high damping. The researchers concluded that their simple screening tool is a good first step to rule out bad ideas, but it cannot replace a full, detailed engineering study before a real plant is built.

The study also addressed a common misconception about nuclear power on islands. Many people assume that smaller reactors are automatically safer because they hold less energy. The researchers showed that this is not true if the reactors share infrastructure. A plant with four small modules that all trip together is just as dangerous as a single large plant. The safety comes entirely from the electrical separation, not the size of the reactor itself. If the modules are not truly independent, the grid sees the same large loss, and the need for emergency backup remains high. The only way to reduce the risk is to ensure that a single event, like a fire or a cable failure, cannot take down more than one module. Furthermore, the researchers highlighted a critical physical constraint: delayed response mechanisms, such as batteries that activate after a short pause, cannot repair an initial frequency-rate violation. If the frequency drops too fast the instant a generator fails, the grid collapses before any delayed backup can kick in, regardless of how much reserve power is eventually available.

In the end, the researchers provided a clear, reproducible method for decision-makers to evaluate nuclear options on islands. They showed that the value of modularity is not infinite; it stops increasing once the modules are small enough to fit within the grid's existing safety margins. They also demonstrated that the financial benefit of this safety is modest, supporting only a small premium in construction costs. For a country like the Dominican Republic, which is considering its energy future, this study offers a way to separate the hype from the math. It suggests that before investing in small modular reactors, planners must first prove that the modules are electrically independent. Without that proof, the promise of a safer, more flexible grid may be an illusion. The study does not say whether nuclear power is the right choice for the island, but it provides a precise tool to measure the cost of safety, ensuring that future decisions are based on clear numbers rather than vague hopes.