The excellent series of articles here in Energypulse.net by Mr. Weissman (1) over the last months have documented well the issues the US faces in our supply of natural gas. His analysis and projections generally match those of the National Petroleum Council’s (NPC) report to Secretary of Energy Abraham (2) – both viewpoints may be succinctly phrased as don’t count on natural gas to provide cheap fuel for future electrical generation. Fields in the lower 48 states are on the decline – an inevitable consequence of resource extraction. Possible significant new supplies inside and along the continental United States are largely locked up due to environmental opposition. Canadian gas is either in the same boat or is expensively distant. Even with a new pipeline to Alaska’s North Slope and deepwater drilling in the Gulf of Mexico, we will need to get into imported liquefied natural gas in a big way just to keep home heating and industrial customers supplied.

The most surprising news to me in the NPC report was the capital cost for LNG infrastructure. To import a benchmark one billion cubic foot per day (bcf/day) of LNG requires a terminal in the country of origin to liquefy the gas, a fleet of tankers, and a terminal on the US coast to re-gasify it and inject into domestic pipelines. All of this, exclusive of gathering fields on the producer’s end and distribution pipelines on our end, is estimated by the sages at NPC to cost $5 to 10 billion.

Let’s take it one step further and estimate the future system costs on LNG-fueled power plants vs. new nuclear power plants.

Given the simplifying financial analysis I’ll be using, we’ll have to estimate the useful lifetime of the LNG delivery system that NPC is proposing. I would guess that the terminal in producing country would have a shorter lifetime than a receiving terminal, given the inevitable depletion of gas but let’s say that the total LNG capital costs can be amortized over 30 years. I’m sure the tankers don’t have a productive life longer than that – I’ve been told in personal communication that an LNG tanker has a design lifetime of 25 years which is rather long for an ocean-going commercial vessel. However, we don’t know since the big ships haven’t been sailing that long. Using 30 years for the whole enterprise seems a reasonable guess - the analysis is not particularly sensitive to lifetime in any case.

Further, the cost of private capital, involving as it does political risks in a foreign country and ships on the high seas, would be relatively steep, say 20%. Receiving terminals in the US have a history of being shut down due to price fluctuations adding volatility to domestic returns. Frankly, setting a hurdle rate for such a massive undertaking is how the Board of Directors of a giant oil company earn their perks. I’m going to use 20% but remember that I have no inside information on the major oil companies’ profit expectations and an actual internal decision analysis would be much more sophisticated.

Since LNG transport is a reasonably mature technology, let’s take the mid-point on NPC’s capital estimate or $7.5 billion. Some projects are going to be cheaper but some are going to be more expensive. A cynical rule of thumb in the US is that the more useful the energy project, the more fierce the opposition. And with the widespread meme amongst the professional worrywarts that each tanker load contains the equivalent energy of a small tactical nuclear weapon, opponents have plenty of ammunition for stirring up the NIMBYs (3). And we can hope that nuclear proponents will not resort to the competitive tactics that the oil interests used against early nuclear power plants. (4)

Putting the above assumptions together into a simple straight-line amortization schedule, just like calculating the payments on a 30 year, fixed rate mortgage, the annual capital carrying cost would be $1.5 billion a year.

It’s going to cost something to keep those sailors on the high seas making the transoceanic shuttle with our fuel. Parts wear out even in LNG liquefaction plants. The harbor masters and the pipe fitters will expect to be paid - we’ll have to add something for operations and maintenance. Until someone has a better guess, let’s say it will cost $150 million a year to operate a one bcf/day system. I’m really playing outside my sandbox here but it wouldn’t matter even if O&M were zero, as you’ll soon see.

The producing countries are going to expect something in return for giving up their gas. It would not be free. I’ve read estimates, maybe wishful, that we can buy plenty of gas at $1.50 per million BTU delivered to the foreign liquefaction terminal. That’s the energy equivalent of $9 a barrel oil. And remember, production costs are no surrogate for prices, only a floor. Our one billion cubic feet a day is going to mean we send someone a check each year for about $560 million. The trend today is for buyers and sellers to prefer spot market pricing for gas, both liquefied and otherwise. Hence, this price will be subject to considerable volatility (5). If the world comes to depend of imported LNG, can an “OLNGEC” (Organization of LNG Exporting Countries) be far behind?

The unit price for landed gas is then about $6 per MMBTU from this set of assumptions. That is somewhat higher than the expectations the gas industry has been promoting of less than $5 gas from LNG yet still within the bounding estimates. (6) If we go with the high end of the capital cost estimate, landed gas looks like $7.45 per MMBTU or the equivalent of over $43 a barrel of oil! Natural gas is good, but is it THAT good?

But we will be needing new, additional combined cycle gas turbine (CCGT) plants to burn this gas. They’ll cost money too, once we’ve supplied the existing merchant plants adequately. If they ran at 100% capacity factor and we use a round number of 10,000 BTU/kW-hr for their heat rate, a billion cubic feet a day would power 4,200 MWe of constant, base load capacity. If we assume that our gas turbines can meet the performance of our nuclear fleet, which is pushing a capacity factor of 90%, then the total installed capacity to use the billion cubic feet a day is about 4700 MWe. At $750 a kW (7), 4,700 MW of combined cycle would cost $3.5 billion. Giving them a break on the cost of capital, we’ll assume that they are in cost-of-service regulation and their capital costs are 10%. We’ll do the same for new nukes latter. Giving the gas turbines a 40 year amortization period seems generous too.

That means that the mortgage payments for new generation hardware will run us $360 million a year. Add in a $100 million for power plant O&M (yet another SWAG but Googlefiable) and the total annual bus bar cost for electricity from running a fleet of CCGTs on one billion cubic feet a day of LNG will cost us about $2.7 billion a year.

Compare that to an equivalent productive capacity in nuclear power plants. A typical evolutionary design nuclear power reactor is about 1500 MW. That ought to cost about $3 billion in capital if we use $2,000 per kW. Given that recent turn-key fixed price bids have been less than that, this includes hefty raises for the nuclear engineers who design and build them (like your author). Amortizing over 40 years at 10%, the mortgage payments would be about $1 billion per year for our 4700 MWe. Of course, the basic hardware is now being designed for a 60 year life with a major overhaul at 40 years but we’ll ignore that. Fuel and O&M should run about $200 million a year per reactor on single sites with some discount for multiple unit sites or another $630 million a year.

Of course, I can hear the screaming now - “What about nuclear waste?” And in a softer, more reasoned tone, “What about decommissioning?” I’ve added the current US Nuclear Waste Fee of one mil per kW-hr sold to my summary of costs. For plant decommissioning, let’s assume a final decommissioning cost of $450 million (8). Reactor licensees use a sinking fund arrangement so I’ll assume a 40 year sinking fund earning a conservative 5%. As you can see from the summary table, both are chump change in the bigger picture.

Bottom line – for the same amount of electricity –

LNG - $2.68 billion a year Nuclear – $1.64 billion a year.

The above is an admittedly simplistic economic analysis but it can not be too far off. For example, taxes are ignored, LNG O&M is a hunch, and non-regulated hurdle rates are only matters of conjecture. Yet, as a matter of national macroeconomic policy, here is a clear business case to prefer nuclear power plants over the LNG-fueled alternative.

Playing with our assumptions a bit, one can force LNG electricity to be roughly equal to nuclear if one assumes the low end of the LNG supply chain capital cost at $5 billion and lower the royalty payments to 50 cents per MMBTU. For perspective on the realism of that latter estimate, the royalties are then just about equal to the cost of nuclear heat IN THE REACTOR. It’s also roughly equivalent to oil at $2.90 a barrel. The spot market might someday see raw feed gas for LNG at that price, but not for long.

We have other choices besides LNG and nuclear. At the delivered prices for LNG, imported oil will be a competitor – if the air quality regulators make the allowances necessary for new oil plants. Combined cycle plants ideally run on natural gas but distillate oil (essentially low quality jet fuel) can be an option albeit at lower efficiency, greater emissions, lower capacity factor, and higher maintenance. The operating and regulatory problems are multiplied if one burns cheaper residual oil. With a current market share of about 2%, it would be a national shame if we were to return to the days circa 1970 when 35% of our electricity came from burning oil. The relative merits of coal-fired generation I’ll leave to others except to say “Not in my backyard!” I can state with confidence that neither oil nor coal would be viable here in my neighborhood (the San Francisco Bay Area) under current regulations.

If I hear another “expert” extol a solar powered future (9), I’m gonna scream. The State of California (who else?) recently built a “demonstration” one megawatt installation, the fifth of its kind and size, at California State University - Hayward. The end result? Electricity at 56 cents a kW-hr (10) averaged over its lifetime excluding the land costs (estimated $4.4 million) and unspecified O&M. The average unit cost for nuclear electricity from our assumptions above is less than 5 cents a kW-hr. Neither price considers capacity credit, for which nuclear is qualified and for which solar is most certainly not. The land dedicated to solar cells required to equal our 1 bcf/day production bogy is over 100 square miles based on the Hayward experience – and Hayward is a fairly sunny place. Can we reasonably expect an order of magnitude improvement in solar costs? I think not.

Of course, the imported LNG-to-electricity scenario is what Enron sold to the Indian government in the Dhabol project, bankrolled in part with a now-defaulted loan of $600 million of US taxpayer moneys from the Import/Export Bank. (11) The stink that project’s failure caused was no bubble in a bathtub.

While my specific numbers are subject to revision and criticism (and tinkering), the take-away conclusion is pretty darn solid – nuclear power will a significantly cheaper source of base load electricity than imported liquefied natural gas. Keep that in mind when you consider our nation’s future energy choices.

References and further information:

1. http://www.energypulse.net/centers/author.cfm?at_id=114
2. http://www.npc.org/ See the Executive Summary, vol.1, page 44, for the capital cost estimate for a reference LNG supply chain.
3. Here’s a sampling of anti-LNG opinion:
   http://www.foepenfro.org.uk/lng.php
   http://www.sepp.org/weekwas/2002/Jan5.htm
   http://www.thebulletin.org/issues/2003/ja03/ja03havens.html
   http://www.greenfutures.org/projects/powerplant/Fay.html (the last is a cogent and dispassionate accident analysis from an MIT professor)
4. Experts, Earthquakes, and Nuclear Power Richard L. Meehan, September 1984, MIT Press
5. For a sanguine view of the LNG market:
   http://www.atimes.com/atimes/Global_Economy/EK11Dj01.html
   For a more tempered view, see http://www.netl.doe.gov/publications/proceedings/02/ngt/Quillen.pdf
6. An optimist’s view of delivery costs is here:
   http://www.energypulse.net/centers/article/article_display.cfm?a_id=235
7. http://www.energypulse.net/centers/article/article_display.cfm?a_id=569
8. http://www.nrc.gov/reactors/decommissioning/funding.html (I’ve used the high end of the government’s estimate.)
9. http://www.energypulse.net/centers/article/article_display.cfm?a_id=580
10. http://www.calstate.edu/newsline/Archive/02-03/030509-Hay.shtml My calculation of the real cost of the CSU-Hayward project is available upon request.
11. http://www.indiatogether.org/interviews/abhay-enron.htm

Author’s note: my spreadsheet file is available via email upon request.