People often discuss the costs of nuclear energy in terms of financing, regulation, or the challenges of “first-of-a-kind” designs. However, a significant part of these costs is simply due to the physical aspects of the designs.
Let’s compare a Pressurized Water Reactor (PWR) and a Boiling Water Reactor (BWR) by examining what is actually constructed.
A PWR incorporates an additional thermodynamic layer, which includes:
Steam generators (multiple large pressure vessels)
A pressurizer with its own systems
High-head primary pumps
Extensive high-pressure piping
All of these components are housed within a large containment structure designed to hold the full primary inventory under pressure.
This results in more components, more interfaces, more supports, and requires more concrete.
In contrast, a BWR simplifies this process:
Water boils directly in the reactor vessel.
Steam is sent directly to the turbine.
It does not use steam generators.
There is no pressurizer.
There is no separate primary loop.
Additionally, BWRs use a pressure-suppression containment rather than a large dry containment structure.
As a result, a BWR comprises fewer major components, shorter piping runs, and requires less material overall.
At some point, costs are not abstract; they are a direct result of complexity.
Of course, there are historical and regulatory reasons why PWRs are more prevalent. But the primary consideration is what actually needs to be built.
When you analyze this, it becomes challenging to understand how adding an entire extra thermodynamic boundary can improve competitiveness.
***
Boiling water reactors were not meant to sit still. They were designed to move.
In a BWR, power control is not something layered on top of the plant. It is built into the core physics.
Generator power is set by adjusting the recirculation pump speed. Increasing the MCP rpm increases the core flow. Higher flow suppresses voids in the fuel channels. With a negative void coefficient, reactivity increases and reactor power increases. Reduce the flow, and the opposite happens.
more rpm → more flow → less void → more reactivity → more power
At the same time, the turbine control system has a much simpler role: keep reactor pressure constant. Steam is taken at saturation, so pressure effectively defines temperature. Hold pressure steady, and the plant's thermodynamic state stays steady.
So you end up with a clean separation:
recirculation flow controls power
turbine controls maintain pressure
No need to “tell” the reactor what to do. It follows through its own feedbacks.
This is why BWRs were originally seen as ideal for load following. Power moves with core hydraulics, while pressure—and thus temperature—remains essentially unchanged.
A design where the physics does the work is often the one that ages the best
***
Most Boiling Water Reactors are overly dependent on the rapid initiation of auxiliary feedwater.
In a BWR, the primary heat sink is the steam path: turbine, condenser, and feedwater returning to the vessel. When that chain is disturbed — loss of condenser, or feedwater interruption — the plant does not settle into a slow transient. It begins to pressurize, and the normal heat sink is effectively gone.
At that point, there is only one way to maintain balance at high pressure: remove steam and replace the lost mass.
And the first part of that happens automatically.
Relief valves will always act. As pressure rises, they open and establish a discharge path to the suppression pool. They become a continuous outlet, effectively a medium-sized LOCA.
The vessel then becomes a one-way system:
core power → steam → relief valves → suppression pool
with no return path.
This is a key transition. The reactor is losing inventory at a steady rate, set by decay heat and system pressure. From that moment on, the transient is governed by mass balance, not by control logic.
That sets a short clock. There is less than an hour to core uncovery if no injection is established. It is not a long, recoverable drift. It is a countdown.
Auxiliary feedwater — or any high-pressure injection — is therefore not just a backup system. It is the only way to close the mass balance at pressure and stop that countdown. And it must do so quickly.
This blurs a boundary that is often presented as clear.
Preventive control and emergency response are no longer separate layers. The moment relief flow becomes continuous, and feedwater is absent, the plant has already crossed from disturbance management into accident progression. The same valves that protect pressure now drive inventory loss. Injection becomes simultaneously the last preventive measure and the first emergency response.
There is no clean handover.
This is where an Isolation Condenser would fundamentally change the situation.
An isolation condenser provides a high-pressure heat sink that does not require feedwater. Steam is routed out, condensed, and returned to the vessel. The mass balance is preserved.
That breaks the one-way behavior.
- Steam removal no longer implies inventory loss
- Pressure can be controlled without continuous relief discharge
- The vessel remains a closed system
Instead of a countdown driven by mass loss, the plant returns to a **heat removal problem with inventory intact**. Time is no longer measured in tens of minutes.
Just as importantly, the boundary between preventive and emergency levels becomes clear again. The system operates in the preventive domain: removing heat, maintaining inventory, and preserving reversibility. There is no immediate need to transition into blowdown and low-pressure injection.
The Isolation Condenser removes the structural dependence on rapid auxiliary feedwater — and with it, the narrow time window that defines the transient in most BWRs.
***
Since you need auxiliary feedwater anyway, it can in principle replace the BWR isolation condenser. That's why both GE and ASEA abandoned the IC after the earliest BWR generations. But that "simplification" happened at the cost of time, and by blurring defense-in-depth
With auxiliary feedwater alone, the plant can be kept under control:
- you inject to maintain level
- you discharge steam to remove heat
Functionally, that works.
But it collapses multiple safety roles into one continuous action. Heat removal, inventory control, and pressure control are no longer separate — they are all tied to the same loop of injection and release. And it cuts the time to core uncovery in case of a failure to inject to only tens of minutes.
That is where the loss shows up.
The isolation condenser keeps functions apart:
- heat removal without discharge
- pressure reduction without inventory loss
- a slower, more reversible transient
It buys time, and it preserves the structure of defense-in-depth.
Without it, auxiliary feedwater forces you into a different regime.
An anticipated operational occurrence — loss of feedwater, loss of condenser, even an SBLOCA — no longer stays clearly in its category. To control it, you deliberately start discharging steam while injecting water.
You are, in effect, creating a controlled, mid-sized LOCA.
Not because the plant failed into it, but because that is the only way to manage heat and inventory at the same time.
That is the blurring.
Preventive action turns into emergency-style management. The boundaries between levels of defense-in-depth are no longer clean. The system still works — but it works by stepping into the next layer early.
Usually, simplification is good, but the IC should not have been abandoned just because the auxiliary feedwater can theoretically do its job. It doesn't do it equally well.
***
In the ASEA-ATOM lineage of Boiling Water Reactor (BWR)s, the transition from a small leak to deliberate depressurization is not left to a single automatic reflex. It is structured as a negotiation between the plant and its operators.
A small leak with an unavailable auxiliary feedwater pushes the plant into an uncomfortable corner. Pressure remains high enough to exclude low-pressure injection, while inventory is slowly being lost. The logic points toward depressurization to enable the Emergency Core Cooling System (ECCS) in its high-capacity modes.
But that same action, taken too early or for the wrong reasons, is irreversible in its consequences.
Depressurization is not just another control action. It is a commitment. Steam is discharged into suppression, margins tied to pressure are given up, and the plant is forced into a low-pressure regime where recovery paths are different and, in some respects, narrower. If high-pressure injection could have been restored, or if the transient had stabilized, the act of blowdown may have made the situation more demanding than it needed to be.
This is why the Automatic Depressurization System is not purely automatic in the simplistic sense. It is armed by conditions—low level, sustained high pressure, lack of effective injection—but its execution is deliberately delayed.
The delay is not a technical limitation. It is a design choice.
It creates a window in which operators can intervene and block the action if they judge that depressurization is premature or unnecessary. The plant prepares to act, signals its intent, and then pauses. In that pause lies the recognition that no fixed logic can fully distinguish between a genuine small-break loss of coolant requiring blowdown and a recoverable disturbance where pressure should be preserved.
The automation is fast enough to protect the plant if operators are incapacitated or the situation deteriorates rapidly, but slow enough to allow human judgment to cancel a wrong trajectory. It does not remove the operator from the loop; it places them precisely at the point where commitment becomes irreversible.
Even a passive blowdown is not universally benign. There are scenarios where holding pressure maintains optionality—keeping high-pressure systems viable, avoiding unnecessary discharge to suppression, and preserving a more controlled thermal-hydraulic state. The plant must therefore avoid locking itself into a single path based on incomplete information.
The delayed depressurization logic reflects a deeper principle.
Control in a BWR is not about maintaining one variable—pressure, level, or flow—but about preserving the ability to choose between competing strategies as the situation evolves. When that choice must finally be made, it should be made consciously.
Blowdown, in this context, is not a failure of the plant.
It is a decision.
And the design ensures that, for a brief but critical moment, that decision remains in human hands.
.***
In discussions about passive safety, the isolation condenser is usually framed as a "decay heat removal system that works without power". That is true, but it undersells its more subtle — and in many situations more valuable — function
The isolation condenser is also a pressure control system.
In a BWR, pressure is not just a thermodynamic variable. It defines what cooling options are available. High pressure locks you into high-pressure injection systems and limits the use of simpler, lower-pressure water sources. That is why traditional accident management often ends up in a forced choice: open relief valves, depressurize quickly, and accept the loss of steam — and with it, inventory.
The isolation condenser changes that logic.
It removes heat by condensing steam in a closed loop and returning the condensate back to the vessel. Steam leaves the reactor, gives up its latent heat in the condenser, and comes back as water. The mass stays inside the system. There is no intentional release.
But because energy is continuously removed, pressure drops.
Not abruptly, not through a discontinuity like a safety relief valve, but gradually and controllably. The system walks the reactor down in pressure while keeping the inventory intact.
That is good in multiple ways:
- You avoid the immediate loss of coolant inventory that comes with blowdown. In long transients, that inventory is your margin.
- You delay or even eliminate the need to transition to low-pressure injection. The system itself creates the conditions under which simpler cooling methods become viable, without forcing the issue.
- You reduce containment challenges. Every kilogram of steam not vented is a kilogram that does not need to be managed later.
- You keep the plant in a more reversible state. Blowdown is a one-way action. Once inventory is lost, you are committed to a different accident path. The isolation condenser preserves options.
Passivity ensures it works when you lose power. But its real value is that it removes the need for aggressive, inventory-destroying actions in the first place. It replaces a step change with a controlled drift.
***
One detail in the ASEA-ATOM Boiling Water Reactors (BWRs) is easy to overlook.
Even after transitioning to internal recirculation pumps, the reactor pressure vessel remains relatively high within the building. At first glance, this seems incongruous.
The answer lies in how the transition was made.
In earlier BWR designs that used external recirculation, a large loop break led to very high blowdown rates. This drove the need for a large suppression pool extending beneath the reactor, thereby increasing the available volume.
With the introduction of internal pumps, the external loop break disappears, along with the necessity for the same level of suppression capacity. The design does adapt:
The wetwell size is reduced.
The space beneath the reactor converts into part of the drywell.
So far, this makes sense.
However, the redesign stops short of a full architectural reset, and the reactor pressure vessel remains relatively high. At the time, this may have appeared to be a legacy issue. Later, it transforms into an asset.
When severe accident management strategies are implemented, the tall lower drywell becomes beneficial. It allows for deeper flooding beneath the vessel during the progression of a core melt.
And depth is crucial.
Deeper water provides a more effective buffer, moderates interactions, and reduces the risk of energetic steam explosions compared to shallow flooding.
What began as an incomplete optimization ultimately evolves into a form of resilience. Not because it was originally intended that way, but because the geometry permits it.
This serves as a valuable reminder: in complex systems, not every advantageous feature is deliberate, and not every legacy constraint is a flaw.
Sometimes, what you did not optimize away is what you end up needing later.