Wednesday, July 22, 2009

The Keys to Lowering Reactor Cost: Investment Costs

Aside from a few minor editorial alterations I have no changes to make to this post.

Reactors are built with borrowed money. Any way you look at it, the cost of money is a major factor in the cost of building reactors. The reactor owner must borrow money to finance the reactor's construction. The borrowing starts even before the first spade of earth is turned, and continues until the current starts flowing to electrical consumers. Since it takes at least 3 years for reactor construction to be completed, and complex reactor projects often take far longer, this means that interest may be accumulating for several years before repayment begins. Thus the cost of interest on borrowed money during the construction phase adds significantly to reactor capital costs.

Let us consider two approaches to reactor manufacture/construction. The first is the traditional approach. The second is the cost lowering approach I advocate.

In the first approach a power company orders a 1 GW Generation 3+ reactor from the manufacturer. Once NRC approval for the project is approved, the manufacturer starts ordering parts and contractors begins site preparations. Money has to be borrowed to pay for these activities and interest charges begins accumulate. Once parts are built they are shipped to the manufacturer for module assembly, and as modules are assembled, they are shipped on to the building site for final assembly. Meanwhile construction activity continues at the site. This goes on for several years. At the end of the construction phase the fuel; is ordered, then loaded into the reactor. Tests are run, and only then does the reactor start generating power. The sale of electricity to the consumers from the reactor produces a stream of money with which to begin repaying interest and principal. We have been borrowing money for 3 years before the first repayment can come in.

The second approach is as follows. The power company orders 10 100 MWe Generation 4 reactors. Their construction is to be spaced over a 3 year period. The factory manufacture approach will allow for rapid assembly of large reactor modules - say a reactor module, a power generating module, and a module for chemical processing units. While the reactor moves down the assembly line site preparation is underway. Once the modules arrive on the prepared site, they are given final assembly. The completed reactor is given its first fuel charge, and after initial testing, electrical production begins. Three months worth of interest has accumulated before the reactor can begin to repay the borrowed money. Then construction of the second unit begins. The small reactor approach, has saved the small reactor owner up too 88% of the accumulated construction phase interest, that would be added to the capital cost of a large reactor project. In addition, during the three year construction project. the owner will see a steadily increasing stream of revenue, which pays not only interest and principle but also contributes to the bottom line of the electrical business.

In addition the decreased risk entailed by the small reactor multiple unit model diminishes investor's risk. Not only is far less money at stake, over a far shorter period of time, but project cancellation due to expense over run or over estimate of consumer demand is fat less likely.

Small, relatively inexpensive, reactors are much more likely to be completed in a timely fashion than big reactors are. Owners are not forced to order more generating capacity than they need as they might with huge one size fits all reactors.

A further observation on reactor construction financing

The current reactor financing system assumes a different set of social goals, than the situation we face demands. The current financial system assumes that the construction of power producing facilities is a speculative investment, whose risk should be born by investors, until the project is complete. Once the production of power begins, the investors are entitled to receive compensation for the risk they task.

This approach leads to the problem we noted earlier, that the accumulation of interest during the project construction phase increases project capitol costs.

Our current social goal is quite different than that assumed by the old regulatory model. Priority needs to be given to the replacement of fossil fuel burning, CO2 producing energy sources, by post-carbon energy sources. This would mean that the sources of the about 75% of American electricity that is currently produced by fossil fuel burning electrical generators must be replaced by post carbon electrical generating sources. No matter what technology is used, the potential cost of greenhouse gas induced global warming far outweighs the cost of changing energy sources, hence the over riding social goal is the change in energy sources, not the question of who should bare the risk. The risk clearly comes from a failure to implement a viable system of financing changes in energy production technology.

Our social goal should be to changing the energy system, to motivate that change by penalizing producers and consumers who do not change, and to see to it that the change can be financed at a reasonable price. Subsidies tend to favor the adoption of energy new technologies, but they may have limited CO2 reduction effect, witness of the peak load inefficiency of California wind generators.

My suggestion would be to link the system of penalties for CO2 generation with a system of rewards for post-carbon energy construction. Hence power companies would have to pay carbon penalties for electricity produced by burning fossil fuels. Rather than going into general revenue accounts however, those penalties would flow into escrow accounts that can be used for for post carbon energy construction. The penalties would be past on to consumers who would be motivated by higher electrical costs use more efficient electrical technology.

Thus the risk of generating new post-carbon electrical generating facilities would be passed on from the investors to the rate payers. It might be complained that this system favors investors over rate payers, but in facto this system lowers rather than raises the costs which rate payers eventually carry. First by insuring that financing for new post-carbon power generating facilities is available, and by lowering the cost of that financing. Thus rate payers are assured that they will have electrical energy when they need it, and that they will not have to carry the cost of construction phase interest once power from the new electrical generating facilities comes on line.


Jim Baerg said...

"My suggestion would be to link the system of penalties for CO2 generation with a system of rewards for post carbon energy construction. Hence power companies would have to pay carbon penalties for electricity produced by burning fossil fuels. Rather than going into general revenue accounts however, those penalties would flow into escrow accounts that can be used for for post carbon energy construction."

Properly done this could reduce a source of opposition to carbon taxes.

In Canada regional differences in the availability of coal & of hydroelectric sites have led to coal being used for much of the electricity generation in Alberta Saskatchewan & Nova Scotia & for a much smaller fraction of the electricity in other provinces. A carbon tax that goes into general revenue would be seen as taxing some regions for the benefit of others.

I would expect similar regional differences to cause similar resentments in other countries.

A carbon tax that pays for building non-fossil generation capacity *near where the CO2 is emitted* would benefit the people who are paying the tax & more closely link the tax to the solution to the problem.

Red Craig said...

Jim Baerg makes a good point. If the government takes money from one group of people and gives it to another, it's a tax, no matter what you call it.

In the US, anything that smells like a tax scheme is doomed at the outset. All three of the major presidential candidates are pushing cap-and-trade with pollution-rights auctions. It's a tax; republicans won't allow it and democrats facing tough challenges won't vote for it. I think the candidates' energy positions have to be discounted for cap-and-trade and for any other features that depend on selling pollution rights.

Charles, your argument for modularizing reactors makes a huge amount of sense. In the past, economy of scale was used to justify bigger plants. As you point out, though, the cost of borrowed capital makes modules look a lot better, especially since the modules can be produced in factory fashion. Getting revenue from some modules while others are being installed offers great appeal.

David Walters said...

For the record, I remain unconvinced that 10 X 100MW reactors/nuclear plants are cheaper than 1 X 1000MW reactor/nuclear plant.

There are somethings that are not necessarily cheaper: the turbine and the generator. A 100MW Brayton cycle turbine will not be that much smaller than a 1000 MW turbine. The same is true with generators. I've seen 100 MW generators and 800 MW ones. The 100 MW generator is not 8 times 'smaller'. A 50 MW brayton cycle turbines is not 1/3 the size of a 150 MW one.

Labor costs and some material costs could almost be the same for both sizes. I'm merely brining this up incase we all start make this assumption. All of Charles points are good it's just they may not all be worked out for all parts of the reactor/power train. We could actually lose money by paying a lot more for the 10 x 100 than a 1 x 1000.

Another area to look at are the very expensive control schemes. Multiple Distributed Control Systems are going to be much more for 10 units than 1. A DCS for a 100MW reactor is going to be similiar to that of a 1000MW reactor...then do we save anything by buying 10 whole DCS systems instead of one? I tend to doubt it.

These costs have to be looked at as well.


Charles Barton said...

David my calculations suggest that an AP-1000 requires 13,000 to 17000 man hours per MW of generating capacity. Labor theory would suggest that the on site reactor construction method do not make efficient use of labor. A reactor uses perhaps 40 tons of steel per MW. At $600 a ton, that would be $24,000. It is quite clear then labor will cost far more than materials, and that most cost savings would come from making more efficient use of labor. Labor economics would almost always favor mass production with automation and robotics, over custom manufacture by craftsmen.

For example, consider the labor economics of mass producing Brayton cycle turbines for 1000 1000 GW turbines verses the mass production of 10000 100 GW units. Automation and robotics save labor cost, but the low number of 1000 GW units may not justify high levels of investment in automated production and robots. The labor saving return on 1000 units are far less likely to justify the investment, than the labor saving per unit on 10,000 turbines. Automation favors the production of large numbers of units.

As for Multiple Distributed Control Systems, would this not be another area for an all out attack on costs by developing new and inovative technologies?

David Walters said...

OK, I think I can articulate the difference then. Certainly on the reactor itself, you are correct.

But, on the other major components, the generator and turbine, I think you may be making an assumption that is not provable, at least not yet. That is the ability to "mass produce" these compenents in the 100MW range. I don't they can be, nor, are they now. They still simply too big and actually more labor, not less, would be required.

Turbine and generator assembly is still a factor floor, but not assembly line, operation. The only advantage your scenerio has in labor savings is that the automation of component creation and assembly can be done silimiarly to automobiles. I think they are still more akin to airplane assembly which is labor intensive.

Right now, if you look at the numbers as they exist...a 100MW turbine is only icrementally cheaper than a 1000MW turbine. They are all assembled by teams.

But what you say about inovation (vis-a-vis the DCS) could apply to this as well, down the road. But right now I don't see it.


Charles Barton said...

David, actually the current LFTR closed cycle gas turbine generator concept calls for 3 turbines per reactor unit. So the individual turbine size would be closer to 30 MWs than 100 MWs. Let us assume that our national goal is to build reactors capable of producing 1000 MWs of electricity. This would require the construction of over 30,000 30+ MW generator turbines. MBA case study question, would you build the 30,000 30+ Mw turbines the same way you would build 1000 1000 MW generators?


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