Mel Brooks, in his classic comedy The Producers, schemed to make money by over-subscribing shares in a sure-to-fail play. Unfortunately for his character, the play became a smash hit, and all the investors wanted their payouts. Since he had sold well over 100% of the interest in the play, he was in a bit of a pickle.
And so it is with natural gas. Clean, easy to use, abundant—natural gas is everyone’s choice for our energy transition away from oil and coal for power generation, industry, homes, and now transportation. Enter oilman-turned-wind-promoter T. Boone Pickens, with a proposal to move U.S. heavy trucks strongly toward natural gas fuel (as compressed natural gas, or CNG). And to enable the offset, the electricity that is currently generated by such gas (about a 21% market share of power generation, according to the Energy Information Administration’s Annuel Energy Outlook 2009, Table 8) would be supplied by new wind farms, built mostly in the Plains States.
The argument is based on simple physical resource reallocation. Large trucks in the United States use roughly 2.8 million barrels of diesel oil per day. If these trucks were converted to natural gas engines over the next few years as the existing stock of trucks turns over, then domestic oil demand would fall by 14% (US DOE-EIA, Petroleum Navigator). Without support or justification, however, this revamped Pickens Plan assumes that all of this demand reduction would come out of imported crude or refined products. The Plan also assumes that CNG will replace diesel on a 1:1 energy equivalence.
Right now, all that gas, and more, is used elsewhere in the economy, especially to generate electricity. Electricity accounts for nearly 30% of gas consumed in the United States. In their base-case projection, the DOE’s Energy Information Administration expects a slight increase in gas use for power generation through 2030, although the fuel declines in importance to about 18% of total electricity production (EIA, Annual Energy Outlook 2009, Table 13).
It seems simple: if wind can supply 20% of U.S. electricity by 2030, then all of that natural gas can be released to fuel large trucks. Supporters also assert that truckers will save (a lot) of money by switching to natural gas. Just recently economist Lawrence Lindsay, writing in the Weekly Standard extolled the virtues of a gas-fueled long distance truck fleet.
At current pump prices of $2.50/gal., diesel for trucks costs about $17.85 per million BTUs. Gas at the Henry Hub in Louisiana is less than one third of that price. In the view of the Pickens Plan proponents, this cost differential is compelling, saving literally billions of dollars annually.
But apples are not being compared to apples, and the conversion argument leaves out important things–like transportation of the gas to market, taxes, compression to fill the trucks, and the station to hold the compressors–factors that are all included in today’s diesel prices. In fact, if one leaves out all of the non-fuel costs and taxes listed above, then at current crude prices, oil costs just $6-7 per million BTUs, not much different from natural gas.
The rub is getting the gas to market. In its recent Annual Energy Outlook, DOE estimated that transmission of gas to market, compression, and taxes equivalent to those levied on diesel will add at least $7 per million BTUs to the price of gas for trucks and other transport users (AEO 2008, Table A13). One more thing, a dedicated large gas truck costs about $75,000 more than a diesel one, with more complex fuel storage. These trucks will also pay some penalty in payload capability since CNG cylinders are heavy. UPS found that an optimized engine used about 10-15% more energy per mile than did a similar diesel vehicle, vitiating most of the remaining fuel price advantage. Overall, operating costs for UPS fleets in Connecticut ranged from about the same as for diesel trucks to 20-30% more.
So if CNG presents no special financial advantage to operators (and this was without equalizing fuel taxes), then reducing oil imports must carry a very large externality to make the Pickens Plan a sensible one. But the externalities do not all move in one direction. If the demand for diesel in the United States falls, then so will its price relative to other refined products, thereby diminishing the putative savings from the fuel switch.
There is an even larger issue, though. The United States currently imports about 15–16 percent of its natural gas, mostly from Canada. About 2–3 percent of our gas supply arrives as LNG. Over the next 20–25 years the DOE expects that gas use in the United States will remain about constant, and that imports will also stay roughly constant. However, most U.S. imports will be LNG by the middle of the next decade. By 2030, the DOE expects that almost 90 percent of gas imports will be LNG (Energy Information, Administration, Annual Energy Outlook 2009, Table 13).
With a growing population and economy, and even with an expectation of no increase in power generation from natural gas, any new gas that goes for large trucks will have to come from somewhere. A full substitution for the diesel used by heavy trucks would require at least 6 tcf/year, more than is used in any other sector of the U.S. economy.
Mel Brooks and LNG
This is where Mel Brooks comes in. Natural gas use is expected to rise only negligibly among residential users due to efficiency improvements—and ditto for industry. And we know that gas molecules cannot be reused, so to meet transportation demand for gas without increasing imports will require that all gas-fired generation be shut down by 2030 and that an additional 1–2 Tcf/d of gas be taken from other end users. But if climate concerns are to reduce the new coal plants currently in the DOE forecast of U.S. generation capacity, and if older coal and nuclear plants reach the ends of their lifetimes and must be retired, then how will the lights stay on?
Easy, they say—just build more gas-fired generation. Everyone is for fast, clean and reliable Plan B for the power sector. But where will the gas come from? LNG, where else. The LNG imports sufficient to replace diesel in large trucks will sum to about 140 million tonnes annually. Apparently, all we will be doing is exchanging gas imports for oil imports. To attract additional LNG supplies the United States will have to start paying more for gas—right now, the production from the various shale formations keeps gas prices relatively low in the United States, making LNG a true marginal fuel. Once LNG becomes a major necessary element in supply (30% to meet the demands of the Pickens Plan), then gas prices will need to rise in order to pull in supplies in a world market where many LNG importers lack domestic supply alternatives. Such a scenario does not sound remotely like the story told by proponents of The Plan.
There is absolutely nothing wrong with CNG vehicles. They are clean, safe and use well-understood technology. Anyone stuck in a Washington, D.C., traffic jam behind a bus certainly appreciates the difference between diesel and CNG exhaust. Moreover, urban fleets may actually be reasonably cost effective conversion targets compared with long distance trucks—the vehicles return to a central garage each night and can use “trickle” compression rather than more costly fast compression, gas infrastructure is already in place—and no government funds are needed. But forcing CNG into inappropriate end uses, using government funds to build pipeline infrastructure in sparsely populated regions of the country, and basing the whole program on vast increases in LNG imports seems to fail the reasonableness test.
 As has been argued on this site repeatedly, a kWh from wind cannot replace one from a central station source on anything approximating a 1:1 basis. Even then, the wind system requires generation backup for a substantial proportion of its operable capacity.
 Total diesel use in the US is currently about 4.5 million b/d, of which about 125,000-300,000 b/d are imported as finished products. The remainder comes from domestic refineries, which use about 65% imported crude oil and feedstocks.
 In the DOE’s Base Case there is substantially increased output from shale formations in the US, which just about covers the fall in output from other sources.
 140 million tonnes of LNG is equivalent to 2.8 million b/d. Currently, the United States imports 12 million tonnes/year and the DOE projects 2030 imports of ~66 million tonnes/year.