Flawed Regulatory Framework: The Nuclear Case

Despite its potential as a powerful tool against climate change, nuclear power has stagnated and struggled to innovate. Constructing new power plants is an excessively lengthy process, and the cost per megawatt-hour is significantly higher than it ought to be.

Interestingly, the root cause of these issues is also a hotly debated topic in the web3 space: regulations. I'm not advocating for an anarchist utopia where everyone does as they please. A clear regulatory framework is crucial to protect users from scams and encourage more institutional actors to enter the scene.

However, a poorly chosen regulatory framework has adversarial effects: at best, it stifles innovation; at worst, it puts more users/citizens in harm. As we shall see, it’s the case of the US nuclear complex.

A good example of a flawed rule used in a regulation framework is the Linear No Threshold (LNT) rule in the Nuclear industry.

This rule was chosen to assess nuclear safety in 1959 and has been used continuously since then. It sets the maximum radiation level to avoid safety issues. All radiation regulations and most radiation casualty analyses are based on this rule. It sounds compelling because less radiation is better, right?

But let’s see how badly crafted this metric is. The rule is based on 3 assumptions:

1. Cell damage is linear in the dose received of radiation as measured in millisieverts (mSv). The more radiation, the more damage to your cells.

   

2. The dose rate (the amount of radiation received in a given period of time) is irrelevant. All that counts is the accumulated dose over time.

3. Mortality and cancer are linear in the amount of cell damage. Similar to the graph above the likelihood increases linearly with the cell damage.

The direct consequences are:

1. Any radiation exposure will never heal.

2. For 1 person, absorbing 5000 mSv in 1h is the same as receiving the same dose over a period of 50 years.

(1) ”Any radiation exposure will never heal.” is basically denying how our body works and how cancer is treated. Cells can repair themselves at a low dose rate. This is fortunate since life emerged 3 billion years ago when the average background radiation dose was about 10mSv a year (about 4 times today’s rate).

In fact, the repair is a bit more complex:

• At a low dose rate, if the cell is damaged, the cell triggers its premature death and avoids deploying useless resources since a great number of other not-damaged cells should be alive and well

• If the dose rate is high enough and the cell's death could create more problems, then the repair process is activated.

• At a higher dose rate, the cell is so damaged that the “repair center” of the cell has a high likelihood of creating errors and mutations, and thus, cancer.

The curve damage resembles something like this (more accurate than the other graph):

Note that the curve is not linear, as the LNT rule hints…

This characteristic is well-known by radiologists to cure cancer. If one can locate the dose so that the edge of the tumor is in the steep part of the curve, one can do much more damage to the tumor than to the rest of the body.

(2) “For 1 person, absorbing 5000 mSv in 1h is the same as receiving the same dose over 50 years” is like saying that being punched approximately once every 3 days for 50 years.

To sum up, LNT is incorrect. Moreover, no data supports it, as described in Why Nuclear Power Has Been a Flop. Nonetheless, the nuclear community has chosen it as the safety north star south star. The standard appears to be more conservative than reality, which seems beneficial for something as critical as nuclear safety.

But let’s see the implications of such a rule.

For a nuclear plant to be below the recommended level based on LNT, it should emit less than 250 μSv per year if radiation leakage should happen. This is equivalent to being exposed to 2500 bananas daily for a year.

Yes, you read that well. Bananas contain natural potassium-40, which emits radiation. I haven’t heard about any cultivator killed by radiating bananas yet. Anyway…

Yes, you read that well. Bananas contain natural potassium-40, which emits radiation. It's like nature's way of giving us a tiny radioactive hug with every bite. I haven't heard of any banana farmers glowing in the dark or developing superpowers yet, but who knows? Maybe we're just not eating enough bananas. Anyway...

This limit (coupled with other factors) pushed regulators to continually increase the requirements for nuclear plants to such an extent that in 1975, the ALARA policy was adopted. ALARA means “As low as reasonably achievable”. Note that there’s no specific requirement; this statement is as fuzzy as a politician's answer to a journalist. With this rule in place, the regulator can drive the cost of the vendor building a plant to what they can afford (while keeping them out of bankruptcy, or else regulators would have no work to do anymore).

Besides that, regulators are not aligned with the construction of any new plant: they share no upside for a successful plant but face all the downsides if something goes wrong. Interestingly, they are incentivized to make the process of giving a license to build and operate a new plant very long and complex. This puts delays in the making and drives prices up. See the huge spike in nuclear plant costs in the 70s (the tech did not change much then).

Due to ALARA, plants must meet new requirements and duplicate all components, including noncritical ones. Thus, adding extra concrete and steel to ensure that emitted radiations in case of leakage stay below 2500 bananas.

This made nuclear completely uncompetitive against other, less environmentally friendly solutions. That’s why we’re stuck building fossil fuel plants that emit tons of CO2 and a lot of waste.

In theory, nuclear should be cheaper than coal and oil. In practice, it’s not.

Another side effect is that by adding unreasonable rules a plant’s design is made a LOT more complex. The more complex, the more likely it is to face unexpected events. This is the case for any critical and complex system, like smart contracts in the blockchain space. You want a smart contract to be damned simple if it’s to hold billions of dollars in a safe way.

Thus, the perverse effect of this rule is to make plants less safe and secure. At odd with the initial plan.

Moreover, the regulation imposes that every new plant design must get a license, but to get a license, you must conduct a test of your design. Tests that you can’t conduct unless you have a license already.

Again, you read that well.

So, one cannot test things before building the actual plant.

This is extremely dumb, knowing that tests in advance are to minimize the number of unexpected problems in production. Take the airspace industry: they don’t just simulate airplane behavior with fancy computers. They actually build several and conduct thousands of tests on real planes before building them at scale. Consequently, a civil plane is one of the safest means of transport on the planet.

Ultimately, it prevents both new and existing players from improving existing designs. Nuclear innovation died.

LNT depicted any low dose as a real hazard and any release, even below the background level of radiation, into an apocalypse. Decisions made in such events are not based on the real impact of radiation. This, in turn, induces unnecessary fear among citizens.

At Fukushima, 2 employees died due to the flooding of the plant. A panic evacuation in the zone was undertaken by local authorities, killing 1600 elderly people because all the hospitals in the region were abandoned. To be perfectly clear, at Fukushima—despite what most newspaper outlets have said—people didn’t die from radiation, but from the tsunami.

For Chernobyl, 220k people were moved. The dose was around 4 mSv, which has no impact on health. From the Health effects of the Chernobyl accident and special health care programs of the World Health Organization:

The largest toll comes from the increase of depression, anxiety, and alcoholism following the event, not the cancer.

Again, if the decisions were made based on better rules following better metrics and the event outcomes were better explained to the public, the consequences would have been greatly reduced.

Sadly, the utterly misleading narrative on nuclear power resulting from the LNT rule deprived us of a very valuable weapon in the fight against climate change.

Apparently, good decisions made on a poor regulation framework result in poor results (sometimes killing people along the way).

As regulatory frameworks for the web3 space are still taking shape, this article serves as a reminder: the rules from today, may backlash tomorrow. The question comes down to the following: do we want to foster innovation while genuinely protecting users in this space, or stifle progress and make it a worse place than what we started with?