Monday, September 22, 2008

The Ultimate Distributive Generator

Rocky Mountain Institute founder Amory Lovins has long advocated distributive electrical generation. There is an extensive discussion of the distributive generation concept in the book "Small is Profitable" parts of which are available in electronic form at the web site. Among the materials found on the "" is a list of 207 Benefits of Distributed Resources. One could go down the list of benefits, and with the exception of benefits that name or apply to renewable generating sources, much of the list would apply to small factory manufactured Liquid Fluoride Thorium Reactors.

The first ten items on the list will be sufficient to demonstrate how well the benefits of LFTR tracks with the benefits of distributive generation:
1 Distributed resources' generally shorter construction period leaves less time for reality to diverge from expectations, thus reducing the probability and hence the financial risk of under- or overbuilding.

2 Distributed resources' smaller unit size also reduces the consequences of such divergence and hence reduces its financial risk.

3 The frequent correlation between distributed resources' shorter lead time and smaller unit size can create a multiplicative, not merely an additive, risk reduction.

4 Shorter lead time further reduces forecasting errors and associated financial risks by reducing errors' amplification with the passage of time.

5 Even if short-lead-time units have lower thermal efficiency, their lower capital and interest costs can often offset the excess carrying charges on idle centralized capacity whose better thermal efficiency is more than offset by high capital cost.

6 Smaller, faster modules can be built on a "pay-as-you-go" basis with less financial strain, reducing the builder's financial risk and hence cost of capital.

7 Centralized capacity additions overshoot demand (absent gross underforecasting or exactly predictable step-function increments of demand) because their inherent "lumpiness" leaves substantial increments of capacity idle until demand can "grow into it." In contrast, smaller units can more exactly match gradual changes in demand without building unnecessary slack capacity ("build-as-you-need"), so their capacity additions are employed incrementally and immediately.

8 Smaller, more modular capacity not only ties up less idle capital (#7), but also does so for a shorter time (because the demand can "grow into" the added capacity sooner), thus reducing the cost of capital per unit of revenue.

9 If distributed resources are becoming cheaper with time, as most are, their small units and short lead times permit those cost reductions to be almost fully captured. This is the inverse of #8: revenue increases there, and cost reductions here, are captured incrementally and immediately by following the demand or cost curves nearly exactly.

10 Using short-lead-time plants reduces the risk of a "death spiral" of rising tariffs and stagnating demand.
Every item on this shortened list would apply to small factory manufactured LFTRs. Thus such LFTR would seem to fit welll into a broad definition of distributive generation. Indeed a strong case can be made, that the small LFTR is an ideal candidate for distributive generator, and that candidates propsed by the RMI carry significant liabilities and limitations. This view directly contradicts the view of the RMI which holds. The RMI favors "Micropower", that is the use of very small decentralized units, but David Bradish has pointed out several problems with this construct in RMI literature. Bradish found that "the largest non-nuclear source of electricity . . . is decentralized generation . . ." Which RMI literature describes as “Non-Biomass Decentralized Co-Generation.” Here my focus diverges from Bradish, who argued that for diverging and conflicting RMI definitions of "Micropower".

My purpose is served by noting that the RMI institute appears by referring to Non-Biomass Co-Generation to be endorsing fossil fuel energy generation, how-be-it in more efficient, decentralized forms. Elsewhere the RMI refers to "Micropower" co-generation as including "turbines and generators in factories or buildings (usually cogenerating useful heat)". As the RMI admits, "Combined-cycle industrial cogeneration and building-scale cogeneration typically burn natural gas, which does emit carbon (though half as much as coal). so they displace somewhat less net carbon than nuclear power could: around 0.7 kilograms of CO2 per kilowatt-hour.(7)"

This is a truly astonishing claim and we ought to expect hard data to back it up. Instead we read in footnote 7 the following words: "7. Since its recovered heat displaces boiler fuel, cogeneration displaces more carbon emissions per kilowatt-hour than a large gas-ï¬ï¿½ red power plant does". That is it, no data at all for what must be seen as an astonishing and highly questionable assertion. But RMI does offer a further argument,
Even though cogeneration displaces less carbon than nuclear does per kilowatt-hour, it displaces more carbon than nuclear does per dollar spent on delivered electricity, because it costs far less. With a net delivered cost per kilowatthour approximately half of nuclear’s, cogeneration delivers twice as many kilowatt-hours per dollar, and therefore displaces around 1.4 kilograms of CO2 for the same cost as displacing 0.9 kilograms of CO2 with nuclear power.

This analysis would and should not go unchallenged, but I will leave the question for others to address. It is clear that the RMI analysis ignores the role that lower cost, factory built small nuclear generating plants can play in distributive generation.

The RMI wavers between viewing renewable micro-generators as supplements to fossil fuel powered central grid generating stations, or as replacements for them. Thus:
68 Distributed resources such as photovoltaics that are well matched to substation peak load can precool the transfomer—even if peak load lasts longer than peak PV output—thus boosting substation capacity, reducing losses, and extending equipment life.

69 In general, interruptions of renewable energy flows due to weather can be predicted earlier and with higher confidence than interruptions of fossil-fueled or nuclear energy flows due to malfunction or other mishap.
Would tend to suggest that some local electricity would be supplied from the grid. Given the RMI's often stated to nuclear power, that electricity could well come from fossil fuel powered generating facilities. The RMI leaves this ambiguous.

It is not without significance that on the "" site we find these words, "Grants from the Shell Foundation, The Energy Foundation, and The Pew Charitable Trusts partially supported the research, editing, production, and marketing of this publication, and are gratefully acknowledged". Shell Oil which is the source of funding for for the Shell foundation, andShell is very much involved in the natural gas business. Shell, while decrying dirty coal is very much involved in coal gasification technology as an adjunct to power production.

If we accept the RMI's view we are forced to acknowledge that electrical generation will continue to produce CO2 for a long time to come, because the RMI does not have a practical plan to rid the Grid of CO2 emitters, and would only somewhat cut back CO2 emissions. Not only does the RMI fall short of demanding the total replacement of CO2 emitting generation facilities, they actually advocate the continued building of new micro-power, natural gas burning co-generation facilities. This would be a problem to those who think that to the extent possible electricity should be generated with no CO2 emissions.

I have a few other observations about the RMI concept of distributive generation. The RMI counts all renewables as distributive generators, but conditions are emerging in Texas and other states in which most of the features of distributive generation appear to be lacking in renewables projects. For example, a recent report from the Electrical Reliability Council of Texas, looked at new grid requirement imposed by the growing West Texas wind industry. The grid expansion turns out to be quite expensive. The report stated:
The estimated costs, excluding collection costs, of the transmission proposal that best meets the criteria for each are:
Scenario 1, Plan A, 12,053 MW, $2.95 billion
Scenario 1, Plan B, 12,053 MW, $3.78 billion
Scenario 2, 18,456 MW, $4.93 billion
Scenario 3, 24,859 MW, $6.38 billion
Scenario 4, 24,419 MW, $5.75 billion.

ERCOT adds:
The cost of transmission is “uplifted to load;” it is rolled into costs that all ratepayers pay (also known as a “postage-stamp” transmission rate because – like stamps – it’s the same access fee no matter where the location is).

The RMI states:
82 Distributed resources have an exceptionally high grid reliability value if they can be sited at or near the customer's premises, thus risking less "electron haul length" where supply could be interrupted.

83 Distributed resources tend to avoid the high voltages and currents and the complex delivery systems that are conducive to grid failures.

101 Distributed resources (always on the demand side and often on the supply side) can largely or wholly avoid every category of grid costs on the margin by being already at or near the customer and hence requiring no further delivery.
Perhaps you have noticed a contradiction between the attributes of distributive generation as suggested by the RMI and the RMI claim that all renewables belong in the category of distributive generation.  I would argue that large renewable projects, located for maximim access to renewable energy rather than proximity to customers, costing billions of dollars to construct, requiring large scale fossil fuel burning backup, and requiring billions of dollars in grid expansion are not distributive generating facilities.  

I will now turn to the question of how the LFTR can be the Ultimate distributive generator.  First, unlike gas co-generators, LFTRs do not burn fossil fuels.  They can be located close to customers.  LFTRs can perform as co-generators.  They can produce both electricity and heat.  There are distinct environmental advantages to nuclear co-generation.   Air pollution becomes a significant issue when fossil fuels or biomass are burned in co-generation facilities.. In addition to CO2. cogeneration produces NOx.    Diesel powered co-generators may also produce SO2.

In contrast, the LFTR produces no air pollutants and no CO2. Heat from the LFTR can be used both for topping and bottoming cycles. Given the use of exotic materials, LFTT could produce heats of 1000 C, and possibly higher. LFTR technology probably should never be pressed beyond 1200 C but PBR technology might provide higher heat, perhaps up to 1600 C. Waste heat from industrial processes, could be run through boilers, for steam generated electrical production.

Topping cycles could use "waste" heat for water or space heating, for lower temperature industrial processes, or for desalinization. The desalinization option would be especially attractive for aired
areas adjacent to sea coasts.

Canadian Reactor Scientist David LeBlanc has proposed a novel LFTR design using an elongated cylinder core. This design would allow a single reactor design to be built with various heat outputs. The only only change would be to the length of the reactor core. Thus factor assembled trasctors can be built to customer output requirements.

Because their higher operating temperature small LFTRs produce electricity with greater thermal efficiency than LWRs. Their high level of inherent safety, and smaller size open unusual siting options, and their high operating temperature will allow them to produce[ost carbon process heat for many heavy industrial processes. Thus small LFTRs posses considerable promise as co-generators. Single LFTR units can be used to produce power for isolated communities, or be placed in compact urban centers to provide space heat and/or hot water for commercial and residential customers. In addition, small LFTRs in the 100 MWe to 300 MWe range, can be clustered into larger power producing units that can generate the equivalent amount of electricity to a very large nuclear plant. Such a facility would have many of the advantages of distributive generation. Units can be built one at a time, lessening the financial risk imposed by the single huge investment approach imposed by the choice of a single huge reactor. The choice several small reactors decreases the effect of reactor down time on grid operations. The choice of LFTRs would of course eliminate down times for reactor refueling. LFTRs could be sited at the location of old coal and natural gas powered generating plants. The LFTR power output cab be matched to the old plant's, thus allowing for simple reuse of the old plants grid hookuo, without modification.

Thus the RMI should recognize that the LFTR fulfills all distributive generation criteria. They don't because it is a nuclear reactor. Not only does the LFTR fulfill the criteria, but it fulfills them better than any of the generating systems proposed by the RMI. It does not burn fossil fuels or require fossil fuel fired backup. It can produce electricity 24 hours a day, 7 days a week, without shutdown for refueling, the onset of night, or changes in the weather. it will be easy to site, will not require dozen of square miles of land to produce electricity, can be cooled with air rather than water. if sea water is chosen for cooling, it can in turn be desalinated. No convention or renewable electrical source can fulfill the distributive generation role better than the LFTR can. The LFTR is thus the ultimate distributive generator.


David Walters said...

Charles, isn't 1600 C a little hot? Doesn't graphite break down at 1400 C?

Also, I've never seen a LFTR 'suggested' design that exceeds 800 c...which is still plenty hot. Just curious...


Charles Barton said...

David, TRISO fuel will tolerate !600 C as will graphite. I am not sure how well they would do with heat and radiation. Molten salt boils at 1400 C, so safety would be a factor above 1200 C in a LFTR. Carbon-carbon composites can tolerate the heat and the molten salt at 1000 C, but may have problems with the radiation. Self healing carbon nanotubes might do the trick.

David Walters said...

Saw this today on carbon nanotubes:

Marcel F. Williams said...

Highly centralized nuclear power production, up to 40GWe, in nuclear parks can take advantage of both the economics of scale and the economics of mass production. If a nuclear park annually adds one or two Gwe of capacity over 20 or 30 years at a centralized location, the overnight capital cost are going to be extremely cheap and the cost of electricity and synfuels from nuclear facilities will be very cheap.

Also, if some of the power at a nuclear park is utilized to produce synthetic peak-load fuels (methanol through water electrolysis and aerocarbon extraction) then small communities far away from nuclear facilities could simply build methanol cogeneration electric turbines or fuel cells to produce electricity and waste heat for the heating and cooling of buildings or for sea water desalinization in coastal areas. The methanol to those communities could simply be pumped in long distance from the centralized nuclear facilities. And carbon neutral nuclear methanol should be competitive in price with natural gas and solar thermal.

However, when there is no increasing demand from regional communities for more methanol fuel, then the methanol can simply be converted into gasoline using Mobil Oil's MTG process since there is no shortage of demand for clean carbon neutral transportation fuel either domestically or foreign.

Marcel Williams

New Papyrus Magazine

The Future of Humanity

Charles Barton said...

The empirical case for economies of scale is weak. Custom manufacture of big reactors constitutes a huge organizational problem, and builders seldom get it right. In the other hand serial production of small reactors, takes far shorter time, leads to a far faster learning curve, more efficient use of labor, cheaper manufacture of parts, and high level of mechanization in production. The infrastructure of your nuclear park would be extremely expensive, and unless it were sea side, it would make large demands on water resources. In addition that much power coming from one source would overwealm the Grid. Massive new Grid infrastructure, costing billions of dollars would be required. Recycling as much of the existing grid structure would be far less expensive. That means matching the power output of old power sources with new.

David Walters said...

I am still inclined to be skeptical on this. Even a 'small' LFTR is still 'big' in asbolute terms. I noted previously that components, such as FD fans, are almost all hand assembled.

You will never get true 'assembly' line built LFTRs...since few commodities of this kind of complexity are really 'assembly line' built. An air plane is a good comparison...many or even most of the components are built assembly-sytle, some even automated, but the actually bringing-it-together is ALL done by hand; in a factory, but by hand. I don't expect the LFTR to be any different.

At some point a comparison on a MW-by-MW study of possible costs in man-hours between an AP1000 and 11 100 MW LFTRs should be done.

PS...I still endorse going toward what Charles suggests. On an empirical test will prove his thesis.

Marcel F. Williams said...

Charles, that depends on how much of the the 40GWe is dedicated towards baseload electricity, how much is dedicated towards the manufacture of on site methanol for peakload electricity, and how much is dedicated for the manufacture of synthetic gasoline, diesel fuel, aviation fuel, methanol, ammonia, and hydrogen.

Nuclear Parks also allow you to economically locate enrichment and reprocessing facilities on site.

Nuclear Parks could also be used as central repositories for all radioactive materials generated within a state.

As far as cooling mechanisms are concerned, cooling ponds combined with dry cooling towers should be able to do the job.

But a nuplex, would grow gradually but continuously, perhaps one or two reactors would go online every year for 20 to 40 years.

However, probably no private capital is going be willing to invest in such a facility which is why I believe that a Federal Nuplex Corporation should be created. The Federal government would simply order power plants like the AP 1000 to be built by private industry at central locations in each state. So private industry would build them, but the federal government would own them.

But it should be dramatically easier and cheaper to gradually build 40 reactors at one central location over a few decades than to build 40 reactors at ten or 20 different locations.

Thanks for the comments!

donb said...

RMI said:
69 In general, interruptions of renewable energy flows due to weather can be predicted earlier and with higher confidence than interruptions of fossil-fueled or nuclear energy flows due to malfunction or other mishap.

This is really comparing apples and oranges. Yes, one can predict that renewable energy flow will be interrupted due to weather. It happens quite often, like almost every day, rather predictably. This is really not comparable to an unplanned outage. The closer comparison would be an interruption of renewable energy flow due to equipment failure even when the weather is favorable.

Any interruption of energy flow when energy is needed is undesirable (even if it is predicted). Most renewable sources have interruptions in spades, so there are many opportunities for interruptions when energy is most needed.

Charles Barton said...

marcel f. williams , with the LFTR "Enrichment" and "reprocessing" are functions of the reactor, and thus do not require separate facilities. Thus part of the deal with LFTR is that with the exception of a start up charge, you simple add more thorium at one end of the reactor and remove Fission products at the other. The reactor does everything else. You still are thinking in terms of the old notion of what nuclear power is. You have to rethink "nuclear power" before you are going to get it. The LWR is part of the problem of nuclear power, not part of the solution. Once you see what the problems are, you might begin to understand the beauty of the LFTR solution.

Warren Heath said...

What Armory Lovins fails to understand ( he is in serious need of tutelage on energy issues), is that his Wind & Desert Solar are the most highly CENTRALIZED energy production that exists today. And his CHP that relies on rapidly diminishing United States NG resources, will require the dangerous import of large quantities of LNG from unreliable Middle East suppliers, which in essence is CENTRALIZED energy production to the extreme. You cannot ignore the extreme remoteness and unreliability of the sources of the fossil fuels that Armory loves so much. No wonder that Armory and Chevron are bosom buddies and partners in the wacky H2 fool cell scam to subvert effort from the four-fold more efficient and 50 fold cheaper electric vehicle. Armory’s Chevron buddies are the guys who buried the large format NiMH battery for Electric vehicles and Solar/Wind type storage applications.

Armory’s ideas are just plain stupid. There is no advantage to distributed power generation which is dependant upon unreliable and rapidly diminishing fossil fuel sources. It makes much more sense to use what’s left of our fossil fuels for petrochemicals and Methonal / DME liquid fuels for use in extreme efficiency engines for transportation – which is more difficult to electrify than simple minded electric building heat & hot water. Besides for hot water, it makes much more sense to use Solar Hot Water, which would way outdo Armory’s distributed CHP plans for economics.

Armory also portrays almost unbelievable ignorance in the cost of Nuclear Power. Sounds like the Hyperion Reactor could be adapted to burn thorium hydride, even hotter with higher fuel burn and higher efficiency, maybe a 15 year refuel cycle, at $1400 per kwe and $500 per kwth, you couldn't beat that with anything on the market today including hydro. They sure could use one of those at any of the big mining sites going up in the NWT, where power costs are running around $0.50 per kwh, burning trucked in fuel oil in diesel CHP generators.


The Hyperion Nuclear Reactor Patent

Explosive Demand for the Hyperion Nuclear Reactor, Armory will put his blinders on for this story

My calculations, using Pickens latest $10 billion Wind project data, put cost at $2,000 per kw-pk wind turbine purchase price = $8,000 per kw delivered plus $1,000 per kw ($2,000 to $3,000 per kw for the Mega-sized National Gore or Pickens Plans) delivered for the extra quadruple over-sized power transmission infrastructure, plus $1200 per kw delivered for the backup NG generator plus $800 per kw installation and road construction etc = $11,000 per kwe or kwth. That will increase substantially if major wind installations on the Gore or Pickens Plan level were to proceed since wind uses 15 to 40 times the raw material of nuclear and 30 to 80 times that of Nuclear CHP. You cannot produce that level of material without causing major shortages and rapid commodity price increases. Current Nuclear Construction costs for LWR's are $2,400 to $4,540 per kwe in the USA. Costs for the LWR's or ACANDU's are projected to drop to ~$1500 per kwe when production supply chains are fully established.

The beginnings of factory Nuclear Reactor construction:

World Nuclear Power predicts up to 1203 GWe nuclear power capacity by 2030 or 35% of total world electricity production. Renewables could not even come remotely close to that, and that calculation neglects the important thermal CHP capabilities of the Nuclear Power and also I believe ignores the development of the new small reactors like the Hyperion and Nuscale.

Armory doesn’t want to hear this: World Nuclear Capacity could reach 1203 GWe by 2030

KLA said...

About the only distributed energy scheme that makes sense is solar thermal. And there in specialized applications.

My wife uses a solar powered clothes dryer (also called cloth-line). This dryer has about the best EROEI of any solar equipment.
Our washing machine is supplied with solar heated water.

Charles Barton said...

Solar water heating and even space heating does well in many localities, but areas with frequent cloudy days are not going to work with any type of solar energy system. It is hard to solar dry clothes on a rainy day,

Kirk Sorensen said...

Warren, the problem with any form of hydride fuel is retaining the hydrogen at elevated temperatures. It quickly evolves out as H2 gas. That's one of the basic reasons that the TRIGA research reactors run at such low temperatures. NASA launched a nuclear reactor (the only time they did it) back in the 1960s that used uranium-zirconium-hydride (TRIGA-type) fuel and it only lasted two months, but its design lifetime wasn't much better, about 18 months before hydrogen evolution from the core reduced reactivity so much that you couldn't maintain criticality.

Hydride fuels are lots of fun at low temperatures but not very useful for high-temperature reactors that need to last very long.

George Carty said...

I think Marcel's concept of nuclear parks would be best used for operation on a plutonium fuel cycle, as this cycle is highly efficient, but has security concerns (although the real menace as far as terrorist nuclear bombs goes would be HEU, not plutonium).

Fewer, larger sites would be easier to defend against attack.

Charles Barton said...

george carty, The thorium fuel cycle in the LFTR is far more efficient than the uranium plutonium fuel cycle, both in terms of neutron economy, and in terms of the cost of fuel recycling.

George Carty said...

Aren't LFTRs at least a decade away though? By the way, could existing reactors be modified to burn thorium?

George Carty said...

I was thinking about some comments by Charlie Stross, where he promoted pebble-bed reactors fuelled by plutonium. I raised the thorium question myself on his blog, but he suggested that thorium reactors were still at least a decade away...

Charles Barton said...

George, LFTRs are at indedd least a decade away. I believe that LFTRs can be the predominate choice to get rid of CO2 from energy generation between 2020 and 2050.

LWRs can operate on a thorium or atleast a partial thorium fuel cycle. But LWRs are not efficient thorium burners. Still thorium cycle LWR would be more efficient reactor than a uranium fule cycles LWR, since U-233 is a mre desirable fuel that Plutonium.

Charles Barton said...

Actually there is on going research in Russia regarding the use of thorium in nuclear fuel. Since PBRs are at least 10 years away from production, 10 years is a a conventional answer. Past research has shown that PBRs do quite well as thorium cycle reactors, but not as well as LFTRs of course.


Blog Archive

Some neat videos

Nuclear Advocacy Webring
Ring Owner: Nuclear is Our Future Site: Nuclear is Our Future
Free Site Ring from Bravenet Free Site Ring from Bravenet Free Site Ring from Bravenet Free Site Ring from Bravenet Free Site Ring from Bravenet
Get Your Free Web Ring
Dr. Joe Bonometti speaking on thorium/LFTR technology at Georgia Tech David LeBlanc on LFTR/MSR technology Robert Hargraves on AIM High