Who will be ThorCon’s EPC contractor?
By Jack Devanney
Engineering, procurement, and construction (EPC) contractors have led recent nuclear plant projects into disasters. Deca-billion dollar cost overruns. Schedules doubling. Some projects being cancelled after squandering billions of dollars. Giant corporations facing bankruptcy. This raises the obvious question: if Westinghouse can’t build a standard nuclear power plant, how in the world will a start up like ThorCon International deliver a nuclear power plant unlike any that has been built before?
Part of the answer is the technology. ThorCon’s power conversion side, the turbine hall and switchgear, are nearly off-the-shelf coal plant equipment. A low pressure fission reactor can be built with conventional metal bending technology. Most importantly, ThorCons will be entirely manufactured in assembly line fashion by a large shipyard, using already developed skills.
Engineering All the big Korean and Japanese yards have basic design and detailed design functions. Basic design takes a potential project far enough to do accurate costing so the yard can bid the job and be confident that it has a good handle on the resources required. At this point, the ship is delineated by about 50 drawings.
After they sign a contract, detailed design takes over. Detailed design not only does the working drawings, but just as important the production scheduling down to per shift detail. This includes scheduling each sub-block and block lift by crane. The weight and center of gravity of each lift is calculated and the lifting lugs are part of the design. Even any scaffolding which will be required in final erection is part of each block design, and installed at the block level. Detailed design and production scheduling cannot be separated.
The yards also have a well tuned monitoring and adjustment process to allow them to quickly respond to hiccups in the production process.
Procurement The shipyards divide purchased material into BFE (Builder Furnished Equipment) and OFE (Owner Furnished Equipment). For a standard ship, there is little or no OFE. Often it is little more than the ship’s stationary and the Owner’s flag. The yard takes responsibility for bidding and buying just about everything from an approved Makers List, which is part of the contract. The yard purchasers are Walmart-like in their ability to play vendors off against each other. The yards are big on-going customers whom no vendor can afford to alienate. And once the contract is signed, every dollar saved goes into the yard’s pocket.
In the case of specialized projects such as a drill ship, the owner may purchase and provide significant portions of the equipment. In the case of the first drill ships, the drilling rig itself was owner furnished. Initially, in ThorCon’s case, OFE may include the Can containing the reactor vessel, moderating graphite, circulating pump, and primary heat exchanger, along with portions of the offgas system. After the yard obtains experience and familiarity with these components, they will become BFE as the drill rigs have.
Construction Construction is the easy part. All the yards have a network of sub-contractors that they use for specialized jobs, and occasionally to level off market ups and downs. But the great bulk of the work is done by the yard itself by a permanent work force, most of whom began their career at the yard and expect to finish it there. They have already received extensive training at yard expense and in most cases worked with the same team for years. No undisciplined boomers need apply. No untrained and undependable locals either.
The Korean yard unions are interesting. They are tough, smart and very disciplined. They know exactly how much money the yard is making and every five years they make sure that they obtain a sizable portion of any gains in productivity. If they don’t, they are perfectly prepared to strike. If they do, the yard is completely shutdown. You do not want to be a scab trying to cross a Korean union picket line. But once the 5 year contract is signed, they live up to it. If a Korean yard worker is not carrying his load, he will hear from both his boss and his union steward. The union wants the yard to make money because they know that will strengthen their hand at the next contract negotiations. The overall effect is that the yards’ labor productivity is more than an order of magnitude higher than on-site construction.
Summary ThorConIsle will rely on the yard for detailed design outside the Can, production scheduling, and much of the equipment purchasing functions. The shipyard will be ThorCon’s EPC contractor.
Jack Devanney is the Chairman of ThorCon International. Jack was a professor of ocean engineering at MIT, who turned to designing the world’s largest (at the time) supertankers, then to this liquid fission power plant planned for Indonesia, using the same block technology. ThorConIsleTM will be constructed by a shipyard on a hull, towed to a shallow water site, then settled to the seabed and powered up.
Thanks for the interesting article. How close is Thorcon to building this? Has a yard provided high level design costs?
I’d also be interested in knowing how the risk of OFE is being mitigated.
Living in an island nation (UK) this could be an interesting development.
I usually get a lot of blow-back when I challenge these MSR designs on another blog. Here goes. There isn’t a snowball’s chance this grandiose plan is going to fly. ThorCon wants to build a fleet without first building a small demonstration plant for testing and validation as if there is some precedent for that. They have major problems.
From the ThorCon offgas system discussion on their web page:
“A critically important advantage of a liquid fuel reactor is that the noble gas fission products, xenon and krypton, are continuously removed…
Each ThorCon module will produce about 0.12 kg/day of Xe and Kr, initially generating a remarkable 1600 kW of decay heat.”
Uh, translation: We have no cladding; all fission products with any vapor pressure are effervesced continuously and any system connected to the primary loop is un-serviceable due to profound contamination. We propose to build a cryogenic offgas treatment facility on-site to deal with this fundamental design problem, which coincidentally is the reason we don’t already have a fleet of these very simple machines.
From the ThorCon recycling discussion on their web page:
“Up to 50 ThorCon plants are supported by a Centralized Recycling Facility (CRF). Normally, Cans are changed out every four years and the fuelsalt is reprocessed every eight years. When the Cans or fuelsalt need replacing, they are shipped to the CRF in a special purpose Canship. The problems of disassembly, decontamination and waste handling are shifted from the plant to this facility.”
Uh, translation: We don’t have any infrastructure to handle the spent fuel salt, but it certainly won’t stay on site forever like the spent nuclear fuel at LWR sites because that is such a political hot potato. We’ll set a precedent for shipping liquid special nuclear material spent fuel in honeypot barges to a secret island shrouded in hand-wavy clouds.
The various MSR groups need to consolidate and push for a modest pilot plant with a cradle to grave plan for no more than several tons of fuel salt. They can build from there after they demonstrate the materials, offgas handling, long term spent fuel storage. If they can demonstrate good handling and effective reprocessing on a laboratory scale, then lobby to bring the tech into production. Nothing else prudent or acceptable IMHO.
I skip over the flowery words and try to address your concerns.
1) Size of the prototype. MSRE was 8MWth. ORNL proposed several different follow-on projects the MOST conserverative of which was the Bettis proposal MSDR. This was a 350MWe demonstration plant that did not attempt to be a breeder, had no extraction of salt-seeking fission products, and was geared to be the most straight-forward next step for ORNL. We are proposing a slightly smaller 250 MWe next step. In many ways our approach is similar to MSDR.
Only one prototype will be built – not a fleet. It will be thoroughly tested prior to getting a commercial license – which is prior to building a fleet. The prototype is large compared to many test reactors – but we already had the 8 MWth version in MSRE so there is little point in repeating that exercise. The safety systems etc. are all sized for the 250MWe power module but initially it will be run subcritical, then very low power, and gradually building up power level as testing proceeds. At each step a technical committee will review the results of the previous tests, and the planned next set of tests to ensure safety.
We considered building a smaller unit but found the cost delta between building a 100 MWth and 557 MWth unit wasn’t that big. If we built a unit that could only go to 100 MWth then we would face the task of scaling up. Normally this is done as a FOAK – which means that the safety testing done at 100 MWth are scaled up by software and analysis and a commercial license is granted to build a full scale system. We feel it is more appropriate to test the full scale system. We can (and will) execute a bunch of tests at 100 MWth but the prototype will be built full scale so that we can test full scale as well.
@Lars Jorgensen
As you know, I am intrigued by many features of MSRs, especially as refined by conceptual designs like ThorCon’s.
Since you point to the Oak Ridge experimental reactors as their proof of concept, it might be worthwhile doing a deep dive on how Oak Ridge experience applies to the off-gas handling and long term fuel storage issues that Scaryjello has raised.
I’m no expert, but I did have an interesting hallway conversation with a guy from Oak Ridge who was involved with trying to create a final disposal plan for the hardened salts left behind by the MSRE. It did not sound as simple as dry storage for LWR fuel, but I’m really interested in learning more.
Off-gas system.
Yes we vent off-gases – as did MSRE. We put the off-gases through a very long delay line (>100 days) by which time the remaining radioactivity is from Kr85. Kr85 is mild enough that it is legal to simply vent it. But we want to reuse our helium so we use cryogenics to separate the xenon from the krypton. The cryogenics are in a part of the system that is serviceable.
Waste handling:
This is a political challenge for every reactor type. Our position is that the critical need is for power and dry cask storage is sufficient for several decades. So that is the initial plan. One day, when the security arrangements have been satisfied we can recycle about 95% of the spent fuelsalt and vitrify the fission products. We have already shown that fluoride salt based fission products (well chemical substitutes anyway) can be vitrified and exceed US standards for vitrification. We have done neutronics simulations to show that we can consume all our own transuranics – once they have been separated. There are several possibilities for the separation process but we couldn’t use it for 20 years at the earliest and it is a politically thorny challenge so we are not working on that now. Dry cask storage works just fine for many decades.
MSRE designers warned that leaving the uranium in the cold fuelsalt was a bad idea. For $50,000 they could remove it – just like they removed the U235 before running on U233. They were not permitted to do this. So the fuelsalt was periodically heated up in the hopes that would prevent any significant release of UF6. After decades they discovered that UF6 had indeed migrated and worry that it might have gotten to the carbon beds and maybe maybe there might be a danger of criticality. We spent gobs of money cleaning ($500M sticks in my mind but I am very uncertain about this). Disposal of MSRE salts had the challenge that they had been left in the drain tank and ignored for many decades without any preparation for long term storage.
For our dry cask storage, we would want to remove the uranium as soon as permitted and in the meantime ensure that the packaging included fluorine absorbers and no path to a moderator.
We expect the dry cask storage will be in solid form and even shipping from the power plant site to the Fuel handling site may well be in solid form. (Note for the prototype these two locations may well be on the same property).
We plan to cool at the power plant site for four years – pretty similar to the minimum time to keep solid spent fuel rods underwater before transferring to dry cask storage.
@Lars Jorgensen
That information is quite helpful and indicates that you have not simply waved hands over the issues. It might be worthwhile to produce a white paper type description of the interim plan for publication on your web site – or mine.
The question that does not appear to be addressed in your response is the handling systems for gaseous fission products. I’m sure you have a plan for those as well that does not involve simply dumping them into the environment – even if they are low or no hazard material in reality. Until further changes are made, we are all still obligated to adhere to ALARA.
The xenon is not radioactive and I expect will have a market value. It can simply be sold.
The krypton will have Kr85 in it. It makes for a good tracer gas to detect leaks so it may well have a market. My guess is though that the output of a few GWe power plants will saturate the market for leak detection gas. So we will have one medium pressure gas bottle per GWe -yr. The half-life is 10.75 years so store the bottle for a century and it isn’t radioactive anymore.
The EPC model itself does not appear to be the problem with the AP1000’s in the U.S. Rather, the implementation of the EPC model by an organization inexperienced with it (i.e., Westinghouse) appears to be the problem. If you do not have your processes, controls, and cost monitoring mechanisms as a second nature in your workforce, the EPC model will destroy you. The recent articles regarding the use of “unlicensed” engineers and the enshrinement of poor designs by Westinghouse is one of the indications of a poor implementation of the EPC model (your engineering has to be right, the first time, almost every time in the EPC model to be successful. This does not appear to be the case for Westinghouse, which fits what I have seen for myself and heard from others regarding the AP1000 fabrication and construction.)
Having experience with both shipyards and EPC contractors, I can attest to the similarity of their processes, though shipyards do seem to have it a bit easier since they have a consistent workforce and do not have to move their physical all of the time. In fact, the best EPC contractors (e.g., the ones that build power plants and other industrial facilities) pride themselves on their ability to manage work forces with varying degrees of skill and capabilities and yet still staying on budget an on schedule while producing a end product that meets the Clients requirements (quality, performance, etc.).
I encourage Mr. Devanney to give the “good” EPC contractors a chance and not just limit their search to shipyards. I also encourage Mr. Devanney to consider US-based EPC contractors and shipyards (there is even an EPC contractor with their own “shipyard” in the US) if ThorCon is not already doing so.
Finally, I do take an issue with the idea that “Construction is the easy part.” I think it is exactly the opposite, construction is the hardest part. It is the hardest part not because it is particularly intellectually challenging, but because it is where your idea becomes reality (and reality has a habit of “getting you” even in the most thought out / planned for everything situations). Plus, Construction is where the big money gets spent and it becomes hard / expensive to fix things that go wrong and where one has to interpret the design, the Codes, and the Regulations while still making a plant that performs to the requirements, is on schedule, and is on budget.
Yes as best I can tell Westinghouse made a series of fundamental mistakes that surprise me. So we can’t take the AP1000s experience in the US as the best we could do using that EPC approach.
However, our target is the developing world and to be cost competitive with coal. Even if the AP1000 was built on budget it is still far too expensive for the developing world.
Our experience in the shipyards come from building four of the largest oil tankers in the world. Each of these is 40% larger than our 500 MWe ThorConIsle. These were built on schedule (1 year) and on budget ($90M). Furthermore, this is normal in that industry. Imagine if you had your power plant building and containment for that price. The steam side is very much like a standard coal plant for which there are many competitive vendors that are shipping in volumes. Yes – the nuclear island will be different and will definitely add to the cost but the nuclear island is relatively small and that is not where the bulk of the costs are.
One challenging area is to keep the regulator focussed just on items critical to public safety and always to minimize the portion of the plant that needs to be the very highest quality. For this reason, software, electricity, turbine/generator, diesel generators, even operators are not critical to our safety case.
Lars,
I am only somewhat familar with ThorCons design, so please forgive me if the answers to my questions are on ThorCon’s website.
The reactors will ultimately go on land, correct? So, you will still need civil engineering activities at the site (clearing, grading, potentially piling, concrete foundations). You will also need to tie into the local infrastructure (e.g., electrical switch yard). Would you use local contractors for that?
How do you plan on transporting the system? I did see on the website that a barge would be used, but what about the “last mile”? Over land transport of large objects can be very hard…it can often drive the size of large (enough) components.
I do appreciate the efficiency and cost containment typical of shipyards. However, there are EPC contractors that are just as good. I keep bringing up this idea because a good EPC contractor could do more than a shipyard since the EPC will move to the plant site and do / control the work there too. The total value equation might be in favor of the mobile builder, even if a mobile builder is not quite as efficient as a shipyard.
Finally, I 100% agree that the regulator needs to only look at the nuclear safety items. Their scope creep (and the industry’s lack of a push back against it) is a bigger problem than most people will publicly admit. We need to break that taboo if we are ever going to actually do something about it.
The reactors will be taken to the site in shallow water (5-10m) and ballasted down to the seabed. So the reactor sits on the ground but right near (or in) the ocean.
The switchyard is contained inside the hull using Gas Insulated Switches.
The tie into the transmission line is the responsibility of the national utility.
We have reviewed the location with them and it works fine for them.
In the future, we do foresee using a mobile contractor and hauling the reactor in 100-500 tonne pieces. Crawler transport can be used to place the reactor within several miles of a large river as is currently done to transport spent reactor cores from the US navy. The reactor would then be assembled on site.
I think the Koreans in UAE recently demonstrated how far you can go within the current LWR system. They got their reactor built at $4/W. So far as I know this was a well executed build and sets a benchmark for what could be achieved within the current paradigm. We are targeting $1/W so we do believe we need to shift how things are done.
On containing safety creep – this is very very important. It has to start with design – so that the safety case can be made despite anything operators, the grid, the turbines, the backup generators, batteries, software etc. do. Yes that means we are not taking credit for all our safety as these things actually do help. BUT we can make the full safety case w/o taking any credit for whatever generators etc do. This then will allow the regulator to focus on just a few items and they can get their job done well in much shorter time. We likely will get international pushback on this (as I have gotten even among leading nuclear engineers so far) but we have to stand firm or king coal will win worldwide.
Reply to Lars Jorgensen’s €1,000/KW
Great article Lars , your target of €1,000/KW is incredibly ambitious but when one considers that the cost of the power generation equipment of a current LWR is ~900//KW ( 15% of overnight capital cost ) and that France built its 58 unit LWR reactor fleet at an average price of €1,800/KW, this does not seem so wildly optimistic – particularly if there is sensible regulation and government is onboard.
Part of the key I think will be ensuring site-specific EPC costs are kept to a minimum .
Diarmuid
Our site specific EPC includes:
a) cold water inlet (maybe 1km to deeper water)
b) breakwaters
c) transmission lines to the site (funded by the utility not us)
d) dredging (depending on the site)
e) perimeter fencing
These are all required for a coal plant as well.
We will likely be required to be more robust than a coal plant in regards
to natural events – in particular seismic. Our seismic system is naturally
fairly robust seismically but in some sites I do expect the foundation and
seismic isolation requirements for us to be more expensive than a coal
plant in the same location. Honestly I think the requirements should be the same but we’ve had enough godzilla movies to train the public otherwise.
G’Day Lars –
I appreciate your pragmatism in compartmentalizing just the nuclear cycle components, (can, pot, pumps and fuel loop etc.) for compliance to any possible nuclear regulations while ensuring that the secondary & tertiary loops together with electricity generation and possible ancillary components such as a Multi-Stage-Flash desalination process (MSF), are external to the nuclear components and in reality no different to those of a gas or coal fired station – which incidentally would never meet nuclear compliance for the radioactive pollution particles.
Which brings me to a couple of queries –
1) Why don’t you choose the more compact Super Critical CO2 Brayton Cycle turbines?
2) Is it feasible to include a MSF desalination plant when clean water remains a significant problem here in Indonesia?.
Finally, having worked in Indonesia as a certified level III ASNT compliance consultant for over 20 years, is there a system to reload the dumped fuel salts back to the can if power or incompetence cause the Helium freeze plug cooling system to fail or will such an event put that particular can out of operation for three or four years?
Cheers – CJ Sazdad
“1) Why don’t you choose the more compact Super Critical CO2 Brayton Cycle turbines?”
At our temperatures the difference in efficiency is negligible – so the difference would be in cost. I do expect the supercritical CO2 turbines to eventually win out over Brayton cycle but they are still in development. So it is a question of risk. We feel we have enough technical risk with the MSR and plenty of political risk with regulatory issues and public acceptance. There just isn’t enough gain going with super-critical CO2 to be worth the additional risk at this stage. As you note, the design separates the turbine from the nuclear side so we should be able to change to super-critical CO2 when they are ready.
“2) Is it feasible to include a MSF desalination plant when clean water remains a significant problem here in Indonesia?”
Yes desal is feasible. To supply a large city with both water and electricity we will use 10-15% of the energy to desalinate the water. But since this is not a universal application and the desalination plant is significant in size we would do this as a companion plant but physically separate.
“Finally, having worked in Indonesia as a certified level III ASNT compliance consultant for over 20 years, is there a system to reload the dumped fuel salts back to the can if power or incompetence cause the Helium freeze plug cooling system to fail or will such an event put that particular can out of operation for three or four years?”
Normal operations involve pumping fuel salt from the drain tank back to the top where it can be directed back to the primary loop. We plan to do this many times in the pre-fission and demonstration plants. A drain and refill will take a plant out of operation for around one day.
Lars