INL's microreactor: A MARVEL machine

Above: An illustration showing the MARVEL reactor as deployed (Source: DOE)

Led by Idaho National Laboratory (INL), the Microreactor Applications Research Validation and EvaLuation (MARVEL) project is a liquid-metal cooled microreactor with a rated peak capacity of 100 kWth. Based on existing technology and using off-the-shelf components where possible, the philosophy behind the MARVEL test bed is to allow for rapid construction to allow demonstrations of microreactor technology and applications promptly.

John Jackson, National Technical Director of the Department of Energy’s Microreactor Program, explains: “MARVEL is a very small microreactor with the primary goal of developing an operational microreactor as rapidly as possible and with as little cost as possible. That means we don’t make technology choices with economics in mind necessarily, or commercialisation in mind, but rather to get something operational as soon as possible that would demonstrate some of the nuanced characteristics of micro reactors.”

The DOE Microreactor Program works synergistically with related DOE programs such as the National Reactor Innovation Center with a goal to accelerate the advanced reactor technologies route to market. The Microreactor Program is designed to support R&D related to the development, demonstration, and deployment of very small, factory-fabricated, transportable reactors to provide power and heat for decentralised generation in civilian, industrial, and defence applications.

As Jackson tells NEI: “We loosely characterised its levelised cost of electricity as something like US$1000/ kWh owing primarily to the fact that we’re not doing this to put commercial power on the grid. We’re doing this to demonstrate capability, build processes and procedures in support of what we expect to be a line of commercial demonstrations to be conducted at the Idaho National Laboratory and other locations.”

Indeed, the design choices were made to get the reactor up and running sooner rather than later rather than with extensive economic considerations. Design choices for MARVEL include the use of four Stiring engines which are expected to generate a total of around 20 kWe and it will be fueled with high-assay, low-enriched uranium (HALEU) fuel of the TRIGA – Training, Research, Isotopes General Atomics – uranium-zirconium hydride design. Although it is rated at 100 kW, MARVEL will be operated at about 85 kWth.

“Stirling engines are more or less off the shelf. We have to do a little bit of radiation hardening for our use, but it’s expected to be minimal. Sodium potassium, NaK, the primary coolant, we’ve got previous operating experience with the Experimental Breeder Reactor and the fuel itself is a very widely used fuel. We did that on purpose because it’s got a very high negative coefficient of reactivity associated with it, so it’s very safe. It also has been used in licensed reactors and it’s very typically used in university reactors. It’s a self-moderating uranium zirconium hydride with slight modification from the university fuel. The slight modification is mostly geometric. The MARVEL fuel element consists of five fuel meats, whereas a standard TRIGA element consists of three fuel meats,” says Jackson.

Fast track to microreactor deployment

With the development goal primarily fast track deployment, operations are expected to begin sometime in the 2025-2026 range. MARVEL itself recently reached a very significant milestone in that it completed its 90% final design at the end of September of 2023. “This is significant because it means that we’re more or less moving from a design phase to an execution phase and we’ve taken a major step toward fabrication and construction,” says Jackson.

The microreactor technology is expected to begin operation at INL’s Transient Reactor Test Facility, also known as TREAT, within the next two to three years.

However, some issues with fabrication have pushed operations back from the original scheduled start date. Jackson explains: “We’re working through some issues with fabricators. There are some small ancillary components like parts of the reactivity control system that have been fabricated but we’ve had a few issues within the supply chain and with fabricators that have kept us from starting actual fabrication of the components. We expect that sometime within the next several months.”

MARVEL is comprised primarily of 316 H stainless steel, a high temperature variant 316 stainless. Jackson explains: “Some of the larger forgings have very long lead times. We are also having trouble identifying a fabricator that can meet all our needs. People haven’t built advanced reactors of this size in the past so there’s a lot of learning associated with it. The supply chain is an issue and resolving typical issues associated with identification of a fabricator that’s suitable and is the best for our needs.”

Construction start is dependent on how quickly INL can resolve the fabrication issues and how quickly the components can be fabricated. Nonetheless, research work has begun on the Primary Coolant Apparatus Test (PCAT), an electric-powered full-scale replica of MARVEL. PCAT will be used to collect data on system’s temperature and coolant flow to validate MARVEL’s modeling and simulation tools. At 4 metres tall and weighing almost a tonne, the PCAT has been loaded with sodium-potassium and lead-bismuth coolants and has been installed at a manufacturing facility at New Freedom in Pennsylvania by Creative Engineers Inc.

Jackson says: “The primary coolant apparatus is the electrically heated surrogate for MARVEL that came online initially on September 19th, 2023. The purpose of the primary coolant apparatus test is not for connection to a micro grid but is more for validation of the MARVEL micro reactor itself by measurement and validation of modelling and simulation tools through the natural circulation of the primary coolant and the secondary coolant demonstration of power conversion.” He continues: “Under initial operation we did measure primary coolant flow and power output but we only took it up to 200 degrees Celsius whereas the reactor will operate at somewhere around 400 to 450 degrees Celsius. The reason we stopped the testing at 200 degrees Celsius had to do with the Stirling engines and their internal coolant. This is a glycol mixture that cools the sterling engines and there was some uncertainty with respect to its cooling capacity and measurement of its flow. So, in order to protect the overall system, we stopped and reassessed and we’re just about ready to restart it and take it up to full temperature and validate the MARVEL modelling and simulation tools.”

Research outcomes

With an anticipated two-year lifespan, MARVEL will offer experimental capabilities that are not currently available at US national laboratories and will allow R&D to be conducted on fundamental features, operations, and behaviours of  microreactors. It will also be connected to the lab’s first nuclear microgrid.

Kurt Meyers, Lead for Microgrids at INL, outlines the microgrid elements of the MARVEL programme: “Our microgrid system that we already have at INL is actually a relocatable deployable microgrid. Our initial thoughts are to deploy that after MARVEL has gone through its initial testing. It already has pretty sizable energy storage with lithium iron phosphate grid-forming inverter systems and a full microgrid control system. We also have renewables, solar and wind, that we can couple with it so we can start to demonstrate some of those controls and couplings between microreactor and other energy input and storage systems.”

Todd Knighton, Lead for Nuclear Applications under INL’s Integrated Energy Systems arm, highlights how MARVEL will potentially tie in with some of the other major research programmes underway at INL: “We want to close the carbon cycle. Our plan and our research with Integrated Energy Systems is to use a nuclear reactor and instead of sequestering carbon use carbon from industry, use it as a chemical building block to produce synthetic fuels and chemicals that could drop-in to the existing economy and infrastructure. We need a large energy input and that’s where nuclear energy comes in. Near term, we’re looking at a lot of heat uses for advanced nuclear reactors. There’s industry out there already that can use that heat and we can substitute advanced nuclear reactors and decarbonise that heat source very easily. Once you get the nuclear actor on the ground, the process of getting to the heat is easy. Nuclear reactor heat can be very competitive on a cost basis with natural gas, especially once carbon carbon credits are taken into account. It could be a very clean and viable direct heat option for industry versus the more costly method of using electricity to provide process heat using solar or wind. That heat from electricity would come at a much higher cost just because of the conversion efficiency from heat to electricity and back to heat. As far as integrated energy systems use of advanced reactor heat, MARVEL is the vanguard system that will enable future advanced nuclear reactor companies to bring their prototypes and demonstrations to INL for future integration with IES such as hydrogen, chemicals, and synthetic fuels production. A co-electrolysis system demonstration by GE is already planned at INL.”

Jackson picks up on the theme of anticipated outcomes from MARVEL: “For the reactor itself, along the themes of supporting accelerated demonstration and deployment of commercial microreactor concepts we’re studying things like start-up transient, zero power, physics testing, and operational characteristics of the microreactor. The start-up procedure will be of very high interest with the neutronics characteristics in the immediate vicinity of the reactor. Some fundamental questions like that will be answered just through operation of the microreactor,” says Jackson. Key research goals of MARVEL include streamlining national lab capabilities by establishing authorisation, qualification, and validation processes for microreactor technologies.

Once operational, MARVEL will also be used to potentially support regulatory approval processes and test microreactor applications for various electrical and non-electric applications, such as thermal storage, water purification, and district heating, as well as demonstrate the capability to manage grid demand and reactor power supply.

MARVEL will also be used to evaluate systems for remote monitoring and operations, and develop autonomous control technologies. Jackson comments: “MARVEL is due to be installed in the Transient Test Reactor facility. It will be more or less remotely operated in that when the Transient Test Reactor is operational people are not in the building per protocol. I assume, similarly, that when the MARVEL microreactor is operational, people will not be in the building so we’ll demonstrate remote operation even though we’re just down the street. Under my base microreactor programme we’re also developing the microreactor Automatic Control System. I am careful to point out that what we’re not doing is seeking autonomy. What we’re trying to do is automate decision-making and action to the extent possible. Automatic or automated control is where through machine learning or artificial intelligence schemes the system in some way makes an automatic decision based on data that it’s receiving. A simple example is if you sense a temperature excursion near the core, you might want to rotate the control drums to a certain degree. That’s a very simple operation that through machine learning we can achieve and we don’t necessarily have to have a human in the loop. We want systems making not binary decisions, but graded decisions based on data input.” Jackson continues: “We’re doing that in the background as a base microreactor programme research and development project in support of supporting all developers, but MARVEL may afford us an opportunity to demonstrate some of that as well.”

Myers also picks up on the connections between MARVEL and other INL projects focused on advanced microreactors: “INL do a lot of work with the military and with the DOD [Department of Defense] and are heavily exploring the options for those mixed remote systems, for island grids that currently are running mostly on diesel. We are really exploring mixtures of nuclear, solar, wind, batteries, desalination type systems and maybe even some central heating and cooling for particular facilities. We are diving pretty deeply into that.”

He continues: “Beyond the Stirling engine type system some of the other advanced reactors that we think are coming from industry are going to have air Brayton power conversion machines on them or steam turbine machines. We’re exploring the characteristics of coupling those with battery storage and grid forming inverter systems and how we can provide a stable electrical system with all those things together and not have to swing the reactor all over the place to cover load steps and have the most controllable and manual system that way.”

Key to this kind of integrated system is the control system that can manage all the different disparate elements. As Myers says: “These particular Stirling engine outputs are relatively easy for us to manage even with our existing microgrid controls but when you move into something much more complex for a power conversion system like an air Brayton cycle or a steam Rankine turbine those, electromechanical interactions between the generator, the loads and the renewables and battery storage are a lot more complicated and so there’s much more coordination and control that needs to be done there.”

Jackson summarises: “The way I like to view MARVEL is as a tiered hierarchy of objectives. First is initial criticality, second is initial operation, and third is sustained operation. Once we address the initial hierarchy of objectives, then we can start to take more risks and deploy more advanced automation schemes, for instance.”

Call for collaboration

Alongside MARVEL’s core research goals, a key aspect of the programme is to support commercial deployment of advanced microreactors. To that end, INL is embracing external partners with a view to leveraging the MARVEL deployment. Knighton observes: “If there are other industries or partnerships that want to demonstrate using MARVEL and the microgrid and even brining their own reactor prototypes and demonstrations, we’re open to those types of possibilities.” Meyers says, “The point is we are not holding this exclusively for our control and utilisation. As an example, when we connect the very small microgrid, companies may want to demonstrate microgrid capability on a nuclear powered microgrid. So, we’re opening the doors, we’re soliciting feedback, and we want companies to pitch their ideas for utilisation of MARVEL and the microgrid.

Knighton continues, “We are planning some demonstrations out on the site and possibly with advanced reactor companies, if people partner with us to put advanced reactors out there. Those could couple with hydrogen electrolysis or with ammonia or synthetic fuels production, so there’s a lot of options that we’re looking at as far as an energy transition powered by nuclear. It’s going to take a lot of energy and we think nuclear is a really great source to provide that massive amount of energy.”

He closes with an open invitation for advanced reactor companies: “We are designing integrated energy systems around advanced nuclear power heat and the heat of existing reactors. We’re looking at a wide variety of reactors that are on the market in the design phase and we’re working with different advanced reactor companies. MARVEL is out front, the vanguard if you will, and is paving the way so that advanced reactor companies can come to INL and we can work with them to do that integrated energy system work of using the enormous and dense heat output to re- power the clean energy economy of chemicals, products, and fuels. I think a lot of that integrated energy system work will happen after we get advanced reactor companies partnered with us and bring some of their demonstrations and prototypes out to INL to work on. MARVEL prepares the way for that to happen.”

Still a few years off deployment, MARVEL is nonetheless intended to lead the way in terms of advanced microreactor commercialisation. As Jackson concludes; “It’s the tip of the spear for some of the long-term objectives for the Idaho National Laboratory.”