On March 8, 2010, the U.S. Department of Energy awarded $40 million to two firms (General Atomics, Westinghouse) for conceptual designs of a high temperature gas cooled reactor (HTGR). The designs are expected to focus on process heat applications for the petro-chemical industry. The reactor is generally known as the ‘Next Generation Nuclear Plant” or NGNP. (Prior coverage on this blog - DOE awards $40 million for NGNP.)The first question is this – once DOE gets the conceptual designs, when will one be built? The second question is what’s the competition likely to be doing while NGNP is coming off the drawing boards?
These comments make no assumption about the type of conceptual design DOE gets for its money. There are lots of alternatives in the GenIV family of ideas. This blog post hits some of the highlights of getting the NGNP to market. It doesn’t dive into the level of detail that would make an engineer happy.
Small reactors out in front
While all this is going on with NGNP, a half dozen or more small reactors are also doing everything in their power to reduce time-to-market. Two of them, the B&W (125 MW) LWR) and NuScale (45MW) LWR have the best chance of getting NRC reactor design certification and inking deals with customers by 2015. This milestone would be achieved five-to-ten years before customers will be able to put their chop on an order for a HTGR reactor.
The two small LWR reactors are targeted at electricity generation for U.S. customer. Deuces, quads, and six-packs are possible combinations for expansion for utilities after they’ve bought the first units.Two other small reactor designs, which are liquid metal cooled nuclear batteries, may achieve market penetration outside the U.S. by 2020 or earlier. These units are targeting so-called “distributed power,” which means the applications are literally off-the-grid at remote locations including military bases, mining camps, and lesser developed countries that simply don’t have the T&D to get the power to customers.
The application mentioned most frequently in the breathless marketing literature of several of the developers of small reactors is to provide process heat for steam in the tar sands region in northern Alberta, Canada. The oil companies there taking heavy crude out of the ground currently burn natural gas to produce the steam needed for the primary extraction process and for primary refining to turn the bitumen into crude oil that can be shipped via pipelines to customers.
The oil companies are understandably skeptical that nuclear reactors can be delivered for their use in a timeframe that makes sense relative to their current business plans. For this reason, it is worth looking at the time-to-market for an HTGR and the major milestones along the way.
Next stop – detailed design
DOE’s press release calls for the two contractors getting the $40 million completing their tasks by August 2010. That’s a very unlikely date and may be a typo in the press release.The reason is that spending $40 million on conceptual design work in five months would represent a new land speed record for spending government money.
Given the number of partners in each contractor’s team, and the technologies they bring to the table, these people may not be able to decide on where to have lunch much less sort out their ideas on what to submit to DOE by August 2010. More likely, the completion due date is August 2011, which makes a lot more sense. I’ve asked DOE about it. I’ll update this part if I get an answer.
Update 03/08/10 7:25 PM MST: DOE is now backing off of the August 2010 completion date. A spokesperson for the agency told this blog late this afternoon, "The Department will now negotiate the final terms and conditions for the awards. Until we have completed negotiations, we won't be able to say definitely when the work will be done."
Assuming August 2011 is the correct date for DOE to get the results of its $40 million in conceptual design studies, someone has to evaluate it. My thought is DOE would do well to have the Idaho National Laboratory (INL) do that work. The INL could evaluate the pros-and-cons of each study, identify gaps, and even provide an overall evaluation on the likelihood the designs could be built with today’s technologies. It would be up to DOE to pick a winner. This way it would be an informed choice.
Update 03/09/10: I've been advised by a reliable source that INL will not be involved in the evaluation of the conceptual design reports. The agency will use independent reviewers.
Once DOE gets the evaluation, which could easily take a year to produce, it is now August 2012. It would take DOE another six months to develop a contract to fund the developer of the successful conceptual design to produce a detailed design. That job could take a couple of years and several hundred million. This puts the project at 2015.
Another five-to-seven years to get an NRC license
Once a detailed design is done, the next step to actually building a reactor is to get the reactor certified by the U.S. Nuclear Regulatory Commission (NRC). Since the NRC has never seen a license application for a high temperature gas-cooled reactor, the review for safety is basically a first-of-a-kind experience for a first-of-a-kind reactor. This is a double dose of “known unknowns.” Clearly, it will take longer than the standard review process for a light water reactor.The time frame here would be about two years to prepare the reactor design certification package. It would take another three-to-five years for the NRC to get their job done. This puts us at 2020 or 2022 depending on how fast all parties in the mix work to achieve results. Of course, this assumes NRC and the vendor get some help from Congress to fund the license review. Otherwise, NRC will be forced to juggle priorities and LWR applications will go first.
Ground breaking in 2025 or later?
Even if the NRC issues a report that certifies the design in 2022, the company that wants to build one still has to apply for a combined construction and operating license. The vendor has the option of submitting the license application in parallel with the design certification which could speed things up. Even so, NRC won’t act on certain parts of the license application until it has wrapped up all regulatory steps to certify the safety of the design.
Here's a short summary list of the milestones based on an optimistic, some would say, "aggressive" schedule. Obviously, your mileage will vary depending on NRC's funding, the complexity of the reactor design, and the levels of government funding. Prolonged U.S. deficits and a slow economic recovery could add delays to this schedule or kill the project. Also, all fuel issues would have to be resolved including testing and fabrication.
- 2011 - complete conceptual designs
- 2012 - complete evaluations and pick a winner, award contract for details design
- 2015 - Complete detailed design including master equipment list and fuel specifications
- 2017 - Submit reactor design certification package to NRC
- 2018 - Submit combined construction operating license application to NRC
- 2020 - NRC issues safety evaluation report on reactor design
- 2021 - NRC issues COL license. award EPC contract, break ground
- 2025 - Hot startup
Most likely, NGNP won’t be built in Idaho. The vendor will want a revenue stream as soon as possible. This means the first plant will be built at a customer site. Most likely such a site would be a major petro-chemical facility like a refinery or chemical manufacturing plant. The objective is to swap out two sets of costs – the cost of crude oil and other fossil fuels and the carbon taxes that surely will be in place by 2012.A significant challenge for the customer will be learning how to operate a first-of-a-kind nuclear reactor in the context of absolutely depending on it for steam. That suggests a breaking-in period of at least several years running the reactor in parallel with existing fossil fuel boilers.
Why process heat first?
The process heat applications for an HTGR would operate at 450-550C. While the temperature inside the reactor could be as high as 850C, the secondary loop would deliver steam at the lower temperature to allow the customer to use conventional materials to harness the heat. To get real value from the reactor in making electricity, experts say it would have to operate at 800-1,000C. The problem is these temperatures pose substantial challenges in terms of the types of materials used in the secondary loop to transfer heat from the reactor core to a turbine.
It makes a lot more sense for a petro-chemical plant to take the process heat application, using temperatures it knows how to control, as well as the steam, with equipment it already owns. it eliminates the need for a whole new round of R&D to develop turbines and heat exchange technologies that would operate reliably at the much higher temperatures to cost-effectively generate electricity.
The process heat niche takes some of the competitive pressure off NGNP since the two LWR designs most likely to get to market in the next five-to-ten years are targeting electricity generation. The size of the NGNP suggests it would not be suited for off-the-grid applications since it would be difficult to transport its components to such sites. This leaves large petro-chemical plants that have both the water access for barges and the need for 600 MW (thermal) of process heat.
Competitive costs?
The best case scenario for payback to process heat customers for a commercial version of the reactor looks like this. Assume a member of the NGNP Alliance burns 1 million barrels of oil/day at $70/barrel. That's a daily cost of $70 million. Every 30 days it burns $2.1 billion in crude oil for process heat and over 300 days it burns $21 billion. DOW chemical, a member of the NGNP Alliance, cited these numbers in briefing slides presented to the Heritage Foundation in 2009.
If a new 300 MW high temperature gas-cooled reactor costs $3,500/Kw, or $1.05 billion, the payback occurs in the first or second year assuming all the oil used for process heat is eventually swapped out for heat from the reactor. The actual payback will be much longer due to the need to amortize R&D, NRC licensing, and start-up costs, which could be an additional $3 billion. Also, the plant would have to reconfigure steam lines and control systems to deliver heat from reactor instead of fossil fueled boilers.
Update 03/09/10: The competitive advantage of the NGNP is its size and affordability. Bruce Power is promoting development of two ACR1000 reactors (1,110 MWe (electric) each) for the tar sands region. At $3,500/MWe, the price would be $3.85 billion each or close to $8 billion. A 300 MWe (electric) NGNP reactor, would be $1 billion as noted in the article. Engineers would like to see more specific financial ratios than the examples used in this blog post.
There are plenty of challenges ahead, and they will take the better part of two decades to resolve them. Readers are encouraged to suggest ways to achieve a shorter time-to-market.
# # #
It makes a lot more sense for a petro-chemical plant to take the process heat application, using temperatures it knows how to control, as well as the steam, with equipment it already owns. it eliminates the need for a whole new round of R&D to develop turbines and heat exchange technologies that would operate reliably at the much higher temperatures to cost-effectively generate electricity.
The process heat niche takes some of the competitive pressure off NGNP since the two LWR designs most likely to get to market in the next five-to-ten years are targeting electricity generation. The size of the NGNP suggests it would not be suited for off-the-grid applications since it would be difficult to transport its components to such sites. This leaves large petro-chemical plants that have both the water access for barges and the need for 600 MW (thermal) of process heat.
Competitive costs?
The best case scenario for payback to process heat customers for a commercial version of the reactor looks like this. Assume a member of the NGNP Alliance burns 1 million barrels of oil/day at $70/barrel. That's a daily cost of $70 million. Every 30 days it burns $2.1 billion in crude oil for process heat and over 300 days it burns $21 billion. DOW chemical, a member of the NGNP Alliance, cited these numbers in briefing slides presented to the Heritage Foundation in 2009.If a new 300 MW high temperature gas-cooled reactor costs $3,500/Kw, or $1.05 billion, the payback occurs in the first or second year assuming all the oil used for process heat is eventually swapped out for heat from the reactor. The actual payback will be much longer due to the need to amortize R&D, NRC licensing, and start-up costs, which could be an additional $3 billion. Also, the plant would have to reconfigure steam lines and control systems to deliver heat from reactor instead of fossil fueled boilers.
Update 03/09/10: The competitive advantage of the NGNP is its size and affordability. Bruce Power is promoting development of two ACR1000 reactors (1,110 MWe (electric) each) for the tar sands region. At $3,500/MWe, the price would be $3.85 billion each or close to $8 billion. A 300 MWe (electric) NGNP reactor, would be $1 billion as noted in the article. Engineers would like to see more specific financial ratios than the examples used in this blog post.
There are plenty of challenges ahead, and they will take the better part of two decades to resolve them. Readers are encouraged to suggest ways to achieve a shorter time-to-market.
# # #
6 comments:
On the subject of small reactors, they simply cannot ever cost effective for several reason. First off, the permitting (environmental permits + COL) cost something like 500 million just to get a piece of paper saying you can start building the reactor. This is going to be the same no matter how big or small the reactor is, so for a 125MW rector the permitting is already the same per MW as a 4 billion dollar full sized reactor. In other words before you pour a drop of concrete or draw a single design drawing your are already at an uneconomical position. The entire idea can only exist if they can get a "mass production" scheme going, but that pretty much means that the first 10 customers have to intentionally pay a billion dollars that they know they can never fully make back on the reactor. Who is going to sign up to go first and take a sure loss?
DOE has another major consideration going, the blue riband commission for disposal of LWR spent fuel. If the design offered uses the spent fuel, it could be a suggestion to the commission and if accepted, result in expeditious clearance by NRC. A simple and economical design of an MSR using TRU's as fissile feed with recovered uranium as fertile fuel for burning up could fulfil the need.
There have been HTGRs licensed in the US. Peach Bottom 1 and Ft. St. Vrain were both HTGRs licensed by the NRC. They did operate at higher temperture and pressure than current LWRs. There is experience out there on the General Atomics plant design.
Instead of simply reserecting the Peach Bottom or Ft.St. Vrain technology, we should move forward. A research team at UC Berkely is studying a pebble-bed derivative called the PB-AHTR. It alsu uses graphite pebble fuel, but would use the superior heat transfer properties of molten fluoride salt to greatly boost the power output and cost effectiveness, as well as reducing the uranium usage by 30%. Unlike the South African pebble bed reactor, the PB-AHTR is quite likely to produce electricity for less cost than conventional LWRs.
Additionally, the PB-AHTR would leverage and advance technology developed previously for the MSBR (which has the fissile fuel desolved in the salt also, to facilitate reprocessing). This would bring us even closer to the MSBR (or LFTR), the most promising Gen IV concept.
Time-to-market could be reduced by lighting a fire in the imaginations of the users of cheap process heat and/or the users of strategic precursor chemicals made cheaper and cleaner by using cheap non-emittiong process heat.
But it really depends on what you mean by "market."
Take hydrogen. If nuclear-powered water splitting came into its own, cheap clean hydrogen would be available in unprecedented quantities. Everybody agrees with this, but few are really pushing for the funding necessary to resolve the outstanding engineering challenges in nuclear water splitting.
And why is this? Because sceanarios for the end-use of all this hydrogen almost all involve hydrogen as a pure fuel for fuel cells or combustion engines. Making THAT a reality would require several Nobel-Prize-class scientific breakthroughs. i.e., the company that develops reliable large-scale hydrogen production from nuclear water splitting has a very good chance of winding up with a great product and no market. So no wonder they're not falling over themselves to make the necessary R&D a priority at the highest levels.
I say we forget about fuel cells and the other end uses that require pure hydrogen. We have the technological knowhow, today, with which we could mix manufactured hydrogen with carbon dioxide and make synthetic hydrocarbon fuels that are identical to the petroleum-based fuels we use today, and whose uptake would require no changes to fuel distribution or end-use technology.
A 450-MW pulverized coal-fired generator running as baseload and capturing 45 percent of its CO2 exhaust could provide the carbonaceous component of 79 million gallons of synthetic gasoline per year.
At a pump price of around $2.30 a gallon (wouldn't that be nice), this gasoline would have a retail value of $118 million a year.
Could you set up a plant to make this stuff for less than $118 million a year?
That kind of question could light the fire in the imaginations of those who are looking to invest in the technologies of tomorrow. And THAT could spark the R&D necessary to quickly make it a reality.
Advantage of PB-AHTR has been mentioned. With the problem of spent fuel storage, a fuel design that is difficult to dispose off has serious limitations. For least amount of radioactive waste, there is nothing to beat fast spectrum or a breeder. MSR or other fluid fuels not only avoid cladding waste but also enable removal of volatile wastes to improve neutron economy and further reduce production of TRU's. I find it difficult to convince myself of the merits of other designs.
Post a Comment