The need to reduce reliance on burning carbon-emitting coal and natural gas for electricity is well recognized on this website. Environmental issues and candidate positions on them get a lot of attention. As to what alternative energy sources should be given the greatest immediate focus, it's safe to say that opinions here are rather divergent.
A lot of debate about nuclear energy addresses such issues as safety, waste storage, and proliferation. But a fundamental question that gets less attention, is whether nuclear plants can even be constructed fast enough. We need to address emissions urgently, both through reducing demand and meeting supply with cleaner alternatives to coal - alternatives that can make an impact within a few years. (In some European nations the electricity generated from wind rose from <1% to almost 10% in just the past few years.)
Recently I've noticed a growing amount of misrepresentation here about how quickly new nuclear reactors could come online in the United States. If you've read the comments here I've seen and gained from them the impression that you could, in say five-six years time, see new American nuclear plants in full commercial operation where there are empty fields right now, ... well, frankly, you've been had. A little mythbusting below the fold...
The misrepresentation about the speed and ease of nuclear plant construction can be seen in recent comments on this site. A sample from a few different authors:
Yes, reactors presently take 10 years to build. But only 3 - 5 years if NIMBYs aren't standing in the way. Wind farms take 5 - 7 years to build if NIMBYs are in the way, and 2-3 years to build if NIMBYs aren't in the way.
[R]eactors can be built within five years and have been by Japan and Canada.
We can't build nuclear plants fast enough... because of NIMBY attitudes, not because of technical or economic constraints.
Were any references given to support those claims? Nope. That's OK, I'll provide some.
How quickly can modern nuclear reactors really be constructed? If they can be up and running in five years, where are the US projects planning commercial operation by 2012 or 2013? As far as I'm aware, current proposals cite 2015-2018 as earliest operation dates. That's 7-8 years with optimism, or 9-10 years with a dose of realism.
Perhaps a look at actual experience with a number of proposed modern designs will shed light on things.
EPR - European Pressurized Reactor
The first EPR under construction is Olkiluoto-3, being built by Areva and Siemens in Finland. It's a big job. From the Areva project website one can learn the following about the Olkiluoto EPR:
- It is "one of Europe's biggest currently [sic] construction projects."
- The peak onsite workforce will be 3000 workers.
- There have been 1500 subcontracts awarded to suppliers from 27 countries.
- Lifting sections into place has required "one of the world's largest heavy-load cranes."
Clearly, building an EPR is a touch more complicated than one would think from reading the comments quoted at the start of this diary.
In terms of technology, the EPR is "the world's first Generation III+ reactor" by Areva's own definition (some outside Areva would disagree) but actually a fairly modest step forward, for example having 37% thermal efficiency compared to 34% for existing reactors.
It's hard to estimate a "normal" EPR build time and cost, as Olkiluoto-3 is around a billion Euro over budget and 18-24 months behind schedule - with major contractor Areva and Finnish owner TVO squabbling over how the extra cost will be divided. If nothing else gets behind, the earliest commercial operation will be at least six years after the start of onsite construction, and eight years from when the Areva-TVO contract was signed. The site of the second EPR (Flamanville, France) had excavation work underway in 2006, with Areva optimistically aiming for completion sometime in 2012.
ABWR - Advanced Boiling Water Reactor
This design has seen operation since 1996 in Japan, and is commonly cited as being fast to build. A frequently repeated statistic is that the quickest two ABWRs (the newest two reactors at the Kashiwazaki-Kariwa plant, where all seven reactors have been offline since the July 2007 earthquake) were "constructed" in 36 and 38 months. This is true, if one takes a very narrow definition of construction that has some use within the nuclear industry but is worse than useless (actually rather misleading) for comparisons with other energy sources. Those 36/38 months were the times from first concrete pour to fuel loading. Using this measure as construction ignores a number of necessary stages, in both (a) after fuel loading before a reactor enters commercial operation, and (b) site preparation before the first concrete pour. For those two ABWR reactors, the time from first concrete pour to commercial operation was 48 and 51 months.
The General Electric "ABWR Plant General Description" document, Chapter 12 (directory of chapters as individual PDF files here), gives a typical schedule for future ABWRs based on the Japanese experience with these times to commercial operation:
- 54 months from the first structural concrete pour;
- 78 months from project confirmation ("authorization to proceed" or ATP);
- 93 months from tender issue.
It's also worth noting that a major factor in the Kashiwazaki-Kariwa ABWR project was extensive use of barges to deliver cranes and large modular components by sea. Construction at a site without similar transport access will require a longer schedule.
ESBWR - Economic Simplified Boiling Water Reactor
This is GE's new design from the ABWR. There are no ESBWRs built yet, but the proposed schedule estimates from GE are very similar to the ABWR. A Department of Energy construction schedule evaluation of several reactor types from 2004 cited vendor schedules for a number of modern designs, with the ESBWR at 60 months - being six months longer than an ABWR (55 months) and equal to a CANDU or Advanced Pressurized Water Reactor. Note that those are "best-case" schedules supplied by vendors, not DOE estimates, and the DOE evaluation highlighted many shortcomings with them - more on this below.
Real world experience with ESBWR construction may not be too far away - there is news this week that Exelon has chosen the ESBWR design for a proposed plant in Texas.
CANDU - Canada Deuterium Uranium
Most experience with CANDU reactors has been in Canada. A significant advantage of CANDU is the ability to use non-enriched fuel. A significant liability has been high maintenance costs. For example, NB Power's plan to spend 18 months and $1.4 billion to refurbish a 680 MW reactor is actually a lot better than other Canadian reactor renovations have cost.
There isn't much recent construction experience in Canada itself as the newest reactor in that country to came online 15 years ago. There are plans in Alberta (where rapid growth in electricity demand is expected, particularly for oilsands operations) to build a new twin reactor (mostly likely CANDU) by 2017.
However, those Canadian CANDU reactors are not a modern design. AECL (Atomic Energy of Canada Ltd) has built modern CANDUs in other nations (China, South Korea, Romania) on time and on budget. The modern CANDU reactors have had a construction method (modular with very heavy lift cranes) and schedule very similar to that for ABWRs above. For example, the two Qinshan Phase III reactors in China (at 728MW each, about half the capacity of an ABWR or EPR) were "built" in 48 and 51 months respectively with a peak labor force of around 7000 workers - making it another example often quoted for the "within five years" myth, but again that is only measuring from structural concrete pour to criticality. The total schedule (from contract signed to commercial operation) was 81 months, and both units were online at around 78 months.
PBR - Pebble Bed Reactors
PBRs offer the potential for improved safety, smaller modular reactor sizes (like, 150 to 300 MW) and no water supply for cooling. Sounds promising. Unfortunately, they seem to be some time away. A leading PBR developer, PBMR of South Africa, plans to start building its first demonstation reactor in 2009 and (if all goes well with that) to begin constructing its first commercial reactor in 2016.
China is also researching small Pebble Bed Reactors, and has a tiny prototype running. There was a some of talk a few years back about China having a full-scale PBR running this decade, but it looks to have been just talk - PBR news from China seems to have gone strangely quiet. So a first commercial pebble bed reactor operating anytime soon isn't sounding very likely.
**** **** ****
It's not just a factor of time. Labor and tools are problems for fission designs too. Nuclear plant construction requires a skilled workforce that is in short supply, and the modular techniques for building modern reactors use very specialized equipment. That 2004 DOE review of schedules for modern reactors that I mentioned earlier (available here, 2MB PDF file) included these concerns:
It is not clear that vendors have fully assessed and planned for all the prerequisites for using advanced construction techniques and specialized equipment (e.g., training, transportation and equipment set-up). Vendors should review the availability of specialized equipment and the transportation requirements needed to achieve construction schedules. Additionally, the requirements for training on these tools and techniques should be assessed. Procurement, transportation, and training needs should be incorporated into pre-construction schedules.
More specifically, in the section Key Vendor Assumptions and Impacts of the Executive Summary:
Each vendor assumed that an unlimited cash flow would be available early in the construction project to support the completion of plant-specific engineering and begin the procurement process for long-lead items. Having these funds readily available at construction inception does not accurately represent the likely U.S. market.
...
Each vendor assumed that an unrestricted labor workforce would be available. This does not accurately represent the U.S. labor market, particularly personnel qualified to perform nuclear safety-related work. While difficult to accomplish prior to site selection, potential owners and constructors should survey labor availability near potential plant sites. This information should then be incorporated into the schedules.
...
Site selections affect the versatility of the vendor project schedules. For example, those that assume only rail and road access are considered more conservative than those that only assume water access. The industry should continue to use caution when evaluating site specific assumptions within schedules.
Further details from the full document:
New construction techniques require the use of specialized equipment. Tools like VHL cranes and automated rebar machines may have to be ordered internationally and well in advance of any physical work. The coordination and availability of these tools may impact the amount of work that can be performed at any one time on the site. Their use will have to be precisely scheduled within the plant and their limited world-wide availability may restrict the number of plants that can be built at one time.
The open-top construction method used by the new plant designs requires a VHL crane to move large components and modules into the buildings. VHL cranes with the capacity, height, and reach required for nuclear plant construction are not currently available in the U.S. and would need to be procured from overseas.
...
The use of new construction technologies to decrease construction schedules requires specialized equipment such as Very Heavy Lift (VHL) cranes and the automated rebar machines to be available early in the construction process. However, there are limited quantities of those tools in the world. Locating, purchasing, and shipping equipment to site may be challenging within the timeframe that is allotted.
In addition to the availability of the equipment itself, labor forces must be trained to use the tools. This is not included in the vendors’ currently identified project scope.
All these factors affect not only the build time at each site, but also the number of sites where concurrent construction can occur.
**** **** ****
Can renewable energy be constructed quickly? It can certainly take a long time when NIMBY opposition and politics gets in the way. A prime example is Cape Wind, which has been arguably the most famous proposed wind energy project in the world for years, yet it may not even get built. However, when the politics and planning processes run smoothly, wind energy can get up and operating very fast. Here are three large U.S. wind farm examples.
Horse Hollow, Texas
This wind farm is, at 735 MW capacity from more than 400 turbines, currently the largest in the USA. It was around 18 months from start to full commercial operation: work began around April 2005, phase 1 operating by late 2005, phase 2 in early 2006, and completion in September 2006.
Peetz Table, Colorado
At 400 MW capacity, the Peetz farm of 267 turbines will briefly be the second largest in the nation. Ground was broken in back in May, the first half was finished and operating in August, and the second half will be complete by the end of the year. That's around seven months, with a peak workforce of 300-350 workers.
Meridian Way, Kansas
This 200 MW capacity project is very new, and just recently announced. Vestas confirmed the order for 67 V90 3MW turbines in early November and the schedule is for site construction to begin late Spring, turbines arriving over the Summer, and project completion by the late 2008.
**** **** ****
Yikes, the diary got longer than I envisaged. There's a lot of information to process, but the situation can be summarized fairly simply:
| Modern nuclear reactor | Modern wind farm |
|
Construction time to commercial operation: | Several years | Several months |
Onsite workforce: | Few thousand | Few hundred |
Onsite lifting equipment: | "Very Heavy Lift" crawler cranes | Regular sized construction cranes |
One final point. There's one other source of electricity emissions reduction, that's a clear winner in terms of speed: conservation. The best electricity generation plants are those that never need to be built. When we reduce our use through inexpensive simple steps the benefit is immediate.
Some energy conservation measures are only applicable during construction or major renovation, and these (while still important) will take years to have a major beneficial impact. Other measures are inexpensive and fast: CFLs for lighting; hot water cylinder wraps; low-flow shower heads; heat-trapping curtains; awnings for summer shade; switching off appliances to avoid standby use; etc etc.
Another inexpensive but powerful measure we have as energy consumers is our choice of supplier. The joules you draw from the grid may have come from a variety of source, and you can't filter out those from coal plants. But you have choices over who gets your money. In every state, there are options (that may or may not require paying a little more) for ensuring the profits from your electricity bill go to renewable suppliers. A great site for information about renewable supply options in each state is the Green Power Network, where you can select your state from this map for green power options and learn about buying it. If you're at all unsure how much of your electricity bill goes to renewable-source suppliers, this resource is a great starting point.