Updated 8/3/2012

RFTS – Renewable Fischer Tropsch Synthesis, and
CARMA – CO₂ Advanced Reforming of Methane Adiabatically.
Storage of renewable energy in standard liquid fuels

For many years, GreenTech leaders have been saying renewables are never going to make a major contribution to our energy needs until we have a better method for storing off-peak clean energy. For even longer, environmentalists and national security analysts have been seeking an alternative to fossil oil for our transportation needs.

RFTS Introduction. For more than five years, we at Doty Windfuels have been working on “RFTS”, for Renewable Fischer Tropsch Synthesis. The process will use off-peak excess wind energy to recycle CO₂ into standard fuels that work seamlessly in the one billion cars and trucks on the road today around the world. The chemistry is fundamentally simple and well understood. First, variable off-peak energy is used to electrolyze water into hydrogen and oxygen; some of the hydrogen is used to reduce CO₂ into CO and H₂O in a catalyzed reaction called Reverse Water Gas Shift (RWGS); the balance of the hydrogen is combined with the CO in a Fischer Tropsch reactor to form a synthetic oil – a mixture of mostly gasoline, jet fuel, and diesel, just like the stuff from petroleum, but with no contaminants. The co-produced oxygen from the electrolysis is a valuable by-product.

Fischer Tropsch chemistry has been well understood for seven decades. During World War II, Germany synthesized some of its fuel by combining CO and H₂ (derived from coal) in an FT reactor. Today, this process is commonly referred to as coal-to-liquids, or CTL. South Africa has the longest history of FTS generation of fuels. Unfortunately, traditional coal-to-liquids production is more environmentally destructive even than tar sands or oil-shale based petroleum because of its enormous on-site release of CO₂. However, if the H₂ is obtained from water and excess carbon-neutral energy, and the CO is derived from CO₂, the fuels are carbon neutral, as we explain in greater detail on other pages on this website.

There is no question that it should be possible to synthesize standard liquid fuels from CO₂ and water using off-peak clean energy. The only step that has not been commercialized is the RWGS process – reducing CO₂ to CO. The question has not been whether such a system would work, but whether it would compete with traditional petroleum derived fuels.
Our analysis concludes that if the correct paths and system design are chosen, overall system efficiency (from input electrical energy to output energy in the liquid fuels) will exceed 58%. If done at a reasonable scale, the synthetic fuels will sometimes compete when oil is as low as $50/bbl – and always when oil is above $95/bbl. After the R&D phase, the equipment capital costs are expected to be low enough to be recovered in 2 to 3 years. Our goal is to develop a process that will allow the world to replace the use of petroleum and tar sands with clean, competitive, carbon-neutral fuels synthesized efficiently from CO₂ and off-peak clean energy.

One of the team’s recent technical papers: Toward Efficient Reduction of CO₂ to CO for renewable fuels. shows that the somewhat related direct-solar-fuels processes face daunting practical barriers because of fundamental laws of thermodynamics governing equilibria in one or more of the required steps. The RFTS process, on the other hand, eliminates such bottlenecks by starting with electrical energy rather than thermal energy or photons to get the needed hydrogen. The technology for electrolysis is mature and efficient. After the electrolysis, there are no highly endothermic reactions, so there are no reactions that are difficult to get to work.

CARMA-GTL – A bridging technology. In spite of the above reasoning, thus far general enthusiasm for Windfuels has been limited – essentially for one valid reason: the current cost of electrolyzers. While analysis shows ROI (return on investment) should be attractive for Windfuels even at current electrolyzer and fuel prices in some regions, the capital outlay for the electrolyzers would be beyond what most investors wish to consider in today’s risk-averse world.

That has motivated us to begin developing a less expensive “bridging” approach to synthesizing fuels from a combination of CO2, methane, and renewable energy. The outcome of this research is a process we have dubbed CARMA-GTL, for Carbon dioxide Advanced Reforming of Methane Adiabatically, with GTL. As explained later, as long as low-cost natural gas is available, our CARMA-GTL process reduces the electrolyzer requirements by a factor of three to ten while actually increasing efficiency.

The Demand for Oil. The strong upward trend of the past three years in the price of oil is bringing gasoline prices back into the media spotlight. We expect to be showing laboratory results by the time the public is again insisting on solutions that show real promise for competing with OPEC.
Windfuels provides the potential for the U.S. to transition from the worlds largest importer of oil to the world’s largest exporter of carbon-neutral transportation fuels.

The company is working on scaling windfuels plant designs down as small as practical with acceptable efficiency and capital cost, since changes in the financial markets argue that the more likely scale-up path would be thousands of 10-MW plants though out the wind corridor rather than a much smaller number of GW plants near large nuclear reactors. Recent simulations are showing 45% efficiency at the 2-MW plant size, 53% at 20 MW, and 58% at 250 MW.

The Availability of Cheap, Off-peak Clean Energy. The “Windfuels” moniker is apt because the vast majority of the inexpensive, clean, off-peak energy that has come online in the past several years in the U.S. has been wind, but nuclear energy may be the best option in some places. A few facts here put the scalability, climate benefit, and expected competitiveness of Windfuels into perspective.

• Approximately 10 TWh (yes, 10 tera-watt-hours) of wind energy was curtailed (idled) in the U.S. last year to keep the off-peak grid energy price from frequently going negative. That is more than the energy in 200 million gallons of gasoline – thrown away. Curtailed wind energy is the U.S. appears likely to continue to grow.

• Economically recoverable wind energy potential in the U.S. exceeds 70 PWh/yr – enough energy to synthesize twice the current U.S. transportation fuel needs and supply all our electrical needs.

• Point-source CO₂ emissions in the U.S. total about 4 Gt/yr – enough to synthesize twice the current domestic transportation fuel usage – about 0.7 Gt/yr. Windfuels are expected to be 85% carbon neutral (fully burdened), while most domestic biofuels are only 5-15% carbon neutral when land-use change is fully considered.

One of the greatest advantages of RFTS is that the electrolyzers would be well positioned to stabilize the grid by ramping up or tamping down in real time to adjust to the amount of excess energy on the grid. This will eliminate negative pricing and curtailment, and ensure that the fuels are completely carbon-neutral (100% of the energy entering the electrolyzers will be matched by additional energy coming onto a grid from otherwise curtailed renewable energy). More information can be found here.

(Note: Much of the above appeared on GreenTechMedia on 2/2/2011,
http://www.greentechmedia.com/articles/read/guest-post-kicking-oil-addiction-permanently-with-windfuels/
and a lively discussion followed.)

Choosing the Best Chemical Paths. It is useful to look briefly, in broad strokes, at some of the reasons why our synthesis choices lead not only to the best option for a renewable fuel today, but also for many decades to come.

There are a number of options for converting CO₂ and water into liquid fuels – and we have looked carefully at all of them. Unfortunately, with the perspective we gained, we concluded that there is nothing that looks like it will make more than a minor percentage of the world's fuels. All may play a part, but there is something (like land use for biofuels) that will limit their ability to scale up to the demand. In our view, only our RFTS (process 7) can be scaled to provide enough transportation fuels.

1. Make ethanol from biomass.
2. Make biodiesel using microalgae and solar energy.
3. Use genetically engineered cyanobacteria and solar energy to make gasoline or butanol.
4. Use extremely concentrated solar energy (at above 1100 K) to make gasoline and diesel from CO₂ and water.
5. Inject renewable hydrogen into conventional CTL plants.
6. Use stranded wind energy to make methanol or ammonia.
7. Develop a novel CARMA-GTL process that is as efficient as possible, scalable from 2-200 MW, that can make all the fuels and chemicals we need from CO₂, shale gas, and water.

Below, we look briefly at the above options, some of which are discussed in more detail in a 2010 ASME-ES paper, “Deployment Prospects for Proposed Sustainable Energy Alternatives in 2020”.

1. Ethanol from biomass. Most of the cellulosic ethanol plants are now failing financially, even though they are using feedstocks that are currently nearly free (sometimes even at negative price), and they are getting a high price plus subsidies for their ethanol. They are generally located in areas where their feedstocks will remain nearly free for at least several years – in some cases even for a decade. However, the only truly competitive use for biomass today is in home heating and co-firing in coal power plants. The demand from these markets exceeds the total potential biomass supply in most regions of the world by factors ranging from 2 to 100. The global average is probably 14. That is, the global demand for home-heating and co-firing exceeds the total potential biomass supply by a factor of 14.

When we already know that the market demand for biomass for home-heating and co-firing could exceed the available supply by a factor of 14 within two decades – and possibly much sooner – it is foolish to advocate making ethanol from a feedstock that will become extremely expensive. Also, using biomass for fuel will inevitably place upward pressure on the price of food. Both of these applications use limited resources on this planet – land, fresh water, nutrients, and human labor.

2. Biofuels from microalgae. This is undoubtedly the most distorted alternative of all alternative fuels options. Related research has been well supported financially for three decades. Our analysis (beginning six years ago, and updated annually) of the various approaches using photosynthetic microalgae has concluded that photosynthetic algae will never produce fuels for under $90/gal – except as a co-product during the treatment of waste water. Some fuels will be able to be made profitably from waste-water (sewage) treatment facilities because the municipalities must pay to have the sewage treated anyway, and it probably doesn’t cost much more to make fuels at the same time. The amount of fuels that can be made profitably from waste-water treatment could supply about 0.5% of the total demand.

A recent DOE-supported study (Lawrence Berkeley National Laboratory) thinks that photosynthetic algae without subsidy from waste-water treatment will eventually be produced for $10/gal. Of course, the authors of this study have all been supported (at least to some extent) by DOE grants for the development of fuels from algae.

The data clearly show otherwise. HRBP spent $22M and showed they could produce fuels from microalgae for $40,000/gal. GreenFuels spent $70M and showed they could produce fuels from algae for $200,000/gal.

Algenol will spend $35M of your tax dollars (DOE plus local support) along over $70M from partners and investors over the next few years. They claim they will produce ethanol from a genetically modified hybrid microorganism at the rate of 100,000 gal/yr from a 40 acre demonstration facility. The upper-level managers of Algenol have a 25-year history of over-stating what they will achieve by factors of 100 to 10,000.

Algenol’s CEO finally let a useful cost number out. They are expecting their photo-bioreactors (PBRs) in large scale production will cost $10M/ha. So if they spend $60M on PBRs, they will be able to afford 6 ha (14.8 acre) of PBRs. The best, credible, short-term, algae production we have see (by Seambiotic) would produce diesel fuels at the rate of 5400 gal/ha/yr in race-track ponds. A reasonable assumption is that long-term production in PBRs might exceed short-term production in ponds by 30%. If the fuel from the micro-organism is ethanol rather than jet fuel, an additional 60% gain in production volume might be possible. So perhaps Algenol will produce ethanol at the rate of 67,000 gal/yr.

If they produce 67,000 gal/yr, and if their bioreactors last 20 years, and if their operating costs are all included in the initial estimate of $100M total, they might be producing ethanol for $127/gal, or $200/gge (gallon gasoline equivalent). However, we suspect their PBRs (which are more fragile and complex than most) will only last 7 years. In which case their ethanol could could cost $256/gal, or $400/gge.

Sapphire has even bigger plans. They will spend over $120M of your tax dollars plus at least $300M of investor’s dollars over the next 4 to 6 years. They might be producing 250,000 gal/yr by 2016 from the initial 400 acres being developed. If their facility then produces steadily with relatively low operating and maintenance costs for 20 years, the fuels they produce will cost about $160/gal. However, with likely O&M costs, their fuels could cost $200/gal.

3. Gasoline from cyanobacteria. The PR campaign of Joule Unlimited made Direct Solar Fuels one of the hottest topics in 2011. They claim they will produce gasoline from genetically engineered cyanobacteria at the rate of 15,000 gal/acre/yr. A peer-reviewed paper with some semblance of scientific objectivity has recently appeared:
http://www.springerlink.com/content/j1414q2u5w25h788/fulltext.html
From this and other materials, it appears reasonable to expect that highly engineered cyanobacteria could be 70% more efficient than microalgae. The theoretical gross solar efficiency limit for bioenergy from cyanobacteria is 12% – the same as the theoretical limit for bioenergy from microalgae. The theoretical limit for motor-grade fuels from microalgae products is probably 7%, as there are losses associated with turning lipids and carbohydrates into motor-grade fuels. Note that gross solar efficiency is defined as the ratio of the energy in the fuels to that in the incident solar energy, and it ignores all the other energy inputs, which thus far have generally exceeded the energy in the fuels produced – often by a factor of 2 to 5. One of the best references here is by Williams and Laurens, 2010,
http://www.rsc.org/delivery/_ArticleLinking/DisplayArticleForFree.cfm?doi=b924978h&JournalCode=EE .
The promoters of Joule Unlimited continually make numerous claims that are simply not supported by any data or valid reasoning. First of all, they claim that 12% efficiency is 7 times the theoretical limit for microalgae. This is plain disinformation. A practical gross solar efficiency has been in the range of 1-3%, but the losses have much less to do with the nature of the hydrocarbons produced than with many other factors that will still limit practical efficiency from cyanobacteria. A few of these cost factors will be reduced a little by cyanobacteria that can directly produce mostly alkanes rather than mostly lipids, triglycerides, proteins, polysaccharides, etc. The problem of separation of the desired products from the culture is also reduced with cyanobacteria compared to most microalgae approaches – though Algenol’s approach is similar.

The PBRs needed by Joule Unlimited to get 15% of the gain they are counting on will probably cost 20% more than those Algenol finds acceptable (where we expect their fuels to cost $200-400/gge). Thus, if there is a 70% inherent theoretical advantage for motor-grade fuels from highly engineered cyanobacteria compared to highly engineered microalgae, plus a (possible) 20% cost savings associated with reduced harvesting costs, a reasonable estimate is that the fuels produced by Joule Unlimited could cost $110-180/gge from a $100M facility. However, it appears their first demonstration will likely be less than 5 acres of the 1200-acre site (that was earlier to be 5000 acres) in Hobbs, New Mexico. Our suspicion is that Joule will go the way of GreenFuels, as all of their publicly released information looks the same – nothing but marketing hyperbole.

4. Fuels from super-concentrated solar heat. Solar Driven Thermo-chemical Conversion (STCC) uses highly concentrated solar energy (much more concentrated than PowerTower CSP) in attempts to produce standard liquid fuels (or just hydrogen) from CO₂, water, and sometimes methane. Government research support (especially in Switzerland, Germany, the U.S., and the UK) has grown rapidly over the past four years, though this is an area that has received very little public and investor attention. STCC is related to Windfuels RFTS in that it seeks to produce standard liquid fuels from CO₂ and water via a syngas (CO+H₂) route.

Some of the processes are referred to as “dry reforming”, carbo-reduction, “solar water splitting”, and many more. In fact, we should point out that there is no standard name, but we refer to all as STCC for convenience.

The best experimental results for STCC thus far have achieved under 2% efficiency (most have been under 0.1%), at enormous costs (over $50/W), and from apparatus having very short useful lifetime – sometimes only a few days. Sadly, these results are as we projected from our theoretical analyses many years ago. Since we have presented more details on the fundamental challenges facing the various STCC processes in a recent publication here (and since STCC appears to be falling out of favor), there is no reason to belabor the point here. We simply note that we believe STCC is much further away from being economically practical than microalgae.

We should emphasize that while there are some similarities between the chemistry of STCC and RFTS, the similarities end there. The economics are worlds apart. If you have any doubts – just follow the above link to get a better appreciation for challenges facing STCC.

For comparison, the Windfuels RFTS processes show 45% efficiency will be practical in a small (2 MW) plant, and 55% efficiency can be expected at 80 MW. Moreover, the cost of the source energy in Windfuels will be much less because it will use cheap, off-peak wind energy and thus allow renewables to grow without any concerns about grid stability. STCC, on the other hand, cannot do anything to improve grid stability.

5. Inject renewable hydrogen into CTL plants. Coal-to-liquids (CTL) plants get the CO (carbon monoxide) needed in the Fischer Tropsch (FT) reactor from partial oxidation of coal. Some of the CO can then be reacted with steam in the water-gas-shift (WGS) reaction to get the hydrogen also needed in the FT reactor. Both of these processes (especially the second) also produce a lot of CO₂, which is vented. If renewable hydrogen – as from wind energy – is used instead of CO followed by WGS, the total amount of CO₂ released is greatly reduced. Not surprisingly, coal supporters are advocating injecting renewable H₂ into CTL plants to make them sound cleaner.

There are problems with this concept. The first problem is that the economics don’t make sense. The biggest part of the cost in CTL is cleanup of the syngas (the CO + H₂) before it goes to the FT reactor. The clean-up process is still required and just as expensive even if some clean H₂ is used during off-peak hours, but now one also must add the electrolyzers and deal with variable rates in the partial oxidation reactor. The economics for using renewable H₂ work only if one completely eliminates the expensive gas cleanup problems – the gas cleanup cost is the largest reason for FT’s reputation of being expensive and not workable at small scale.

6. Wind methanol, DME, and ammonia. The company Blue Fuel Energy http://www.bluefuelenergy.com/index.html , may soon begin a rather large scale demonstration of methanol production from CO₂ and water using stranded wind energy. This is clearly a much more competitive and scalable option than any of the preceding options (other than Windfuels). They expect to pay about $55/MWh for the energy from an exceptionally windy ridge.

Some researchers have said that methanol production can work efficiently with the right catalysts starting with just CO₂ and H₂ in the syngas, but commercial methanol production processes always begin with a mixture of CO₂+CO+ H₂ for the syngas, and (wisely) that is what Blue Fuel Energy plans to use. They will get the needed CO by sending some of the CO₂ and H₂ through an RWGS reactor.
The methanol can then easily be made into dimethyl ether (DME), and the DME can easily be made into gasoline, though there are efficiency losses of at least 5% for each step along the way. The company plans to sell all three products – methanol, DME, and gasoline.

The problem with this plan lies in the economics. Blue Fuel Energy plans to operate from largely stranded wind – wind with little other market. Our model for Windfuels, on the other hand, is to operate primarily from off-peak wind – which is less than one-third as expensive because the wind farms are largely supported by the high prices they get for their energy during peak-demand hours.

Many others (most famously, the late Mat Simmons) have advocated making ammonia from stranded wind energy and air. The process is straight forward, but again the problem lies in the economics. Ammonia from stranded natural gas is much cheaper, and will remain cheap for decades. It is true that ammonia could be used as a motor fuel, but ammonia is much more hazardous and less efficient than hydrocarbons or ethanol, so ammonia will never be accepted as a motor fuel. There is a strong and growing need for ammonia for fertilizers, but that need will be met by natural gas and coal for decades.

Returning to Blue Fuel Energy, we expect they will not make money from carbon-neutral methanol, as methanol from coal is much cheaper per unit energy than any other liquid fuel – usually about half as expensive as diesel. Demand for methanol blending into gasoline will be limited (as that makes the gasoline more toxic and less suited for hot climates), and there is also not likely to be much demand for carbon-neutral DME. DME has numerous drawbacks as a motor fuel compared to gasoline, ethanol, diesel, and propane. We suspect most of their methanol will be converted to gasoline, for which there is a strong market. The price per energy for jet fuel and diesel is stronger than for gasoline, but a simple process for efficiently making jet fuel or diesel from methanol, DME, or gasoline has not been developed.

7. RFTS – Make jet fuel and diesel from off-peak wind.
This is where we started at the top of this page, and we will not repeat those comments here, but rather go a little further into the technical justification of RFTS.

One advantage of RFTS is that it is extremely flexible. If market conditions change, it will be relatively easy to change the primary product emphasis from jet fuel to gasoline or diesel.

The primary basis for the sentiment against FTS is the perception that it is capital intensive, and that it cannot be made competitive at small scale – because there are too many complicated sub-processes. Indeed, that’s what Shell, BP, Sasol, and other big oil companies would have you believe – because they don’t want a small company to even think of trying to compete. (But you probably already know that you can’t believe everything big oil says.)

There has also been considerable sentiment against FTS on the basis that the reaction products are not highly selected, so much more effort is required in the separations than for methanol synthesis, for example. However, the advances in our patents and proprietary technology now make that a non-issue.

It is important to appreciate how much easier it will be to scale down a Windfuels/CARMA plant than a conventional FTS plant making diesel from coal or natural gas. The biggest single factor driving conventional GTL plants toward large scale is that it has been very expensive to get the high-purity oxygen they need for partial oxidation in the syngas production at a plant size below 30,000 bbl/day. This, of course, is not needed in the RFTS plant (we don’t need high purity oxygen – we produce it). The other major drivers toward large plants have been the costs of syngas clean-up and product upgrading. Both of these are now greatly simplified. The CO2 coming into the RFTS plant will be three orders of magnitude cleaner than coal, and its final cleanup is much simpler for other reasons too. Elsewhere on this website some of the innovations that further help are discussed in more detail – but naturally, some are proprietary
.
The arguments against biomass to liquids (BTL), as advocated by Enerkem and Rentech, are even more compelling:

(1) Biomass has half the energy density of coal, so transportation and initial processing costs are doubled;
(2) Biomass has more impurities than coal, and of greater unpredictability, so syngas cleanup costs are 50% greater from biomass than from coal;
(3) BTL plants are always much smaller, because the feedstocks are not available nearby in sufficient supply for large BTL plants;
(4) The BTL feedstocks are usually much more expensive after the first few years;
(5) The BTL companies are usually not motivated to make the process competitive – they’re primarily motivated by the value of the initial PR.
(6) The waste stream from BTL is greater than from CTL.

Of course, when the feedstock is at negative price, as it sometimes is, BTL should work. But there is only the potential for enough negative-priced biomass to supply 0.5% of our demand for liquid fuels.

Returning to RFTS, it is important to emphasize that the FTS reactor, when properly designed, can be scaled down by more than four orders of magnitude from typical commercial CTL or GTL sizes with little effect on cost effectiveness, general design, or engineering parameters. (The diameter of the tubes used in the FTS reactor will be the same whether the fuel production is 3 gal/day or 30,000 bbl/day.) The same is true of most of the other major plant components except the waste heat conversion, so efficiency suffers a little at smaller scale. Still, our detailed simulations are showing 33% efficiency at the 200 kW size, 45% efficiency at the 2-MW plant size, 53% at 20 MW, and 58% at 250 MW.

The other major difference between our approach and that of some other groups advocating “closing the carbon cycle” (making fuels from CO₂) is that we have concluded the best routes will all begin by reducing CO₂ to CO in an RWGS or CARMA process. Others have said that it would be better to instead put several decades of work into developing catalysts that might make it possible to get the desired level of conversion and selectivity with the CO₂ and H₂ sent directly into an FT reactor. There have been a number of laboratory experiments suggesting this might eventually be possible; but if the slow progress over the past decade is a valid indicator, it is many decades away. The Windfuels RWGS-FTS and CARMA processes, on the other hand, will work today.

Making an RFTS plant work optimally requires numerous changes compared to a fossil-diesel FTS plant. Not surprisingly, many of the features disclosed in detail in the pending Doty patents are somewhat at odds with current trends in the design of fossil-diesel plants (most notably, the primary separations processes, waste heat utilization methods, the reactor design, and operating conditions). However, the bottom-line results from the simulations speak for themselves. The combination of the various innovations permits a factor of two increase in efficiency compared to what most have expected, and capital costs will be less than those for a cellulosic ethanol plant of similar capacity.

--

Iceland’s CO2-to-methanol plant:
http://www.carbonrecycling.is/index.php?option=com_content&view=article&id=14&Itemid=8&lang=en

Key References:

1. S Phillips, A Aden, J Jechura, and D Dayton, “Thermochemical Ethanol via Indirect Gasification and Mixed Alcohol Synthesis of Lignocellulosic Biomass”, NREL/TP-510-41168, 2007. http://www.nrel.gov/docs/fy07osti/41168.pdf

2. M Xiang, D Li, H Qi, W Li, B Zhong, Y Sun, “Mixed alcohols synthesis from CO hydrogenation over K-promoted ?-Mo2C catalysts”, Fuel 86, 1298-1303, 2007.

3. JE Whitlow and C Parrish, “Operation, Modeling and Analysis of the Reverse Water Gas Shift Process”, 2001 NASA/ASEE summer program, JFK Space Center, http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20020050609_2002079590.pdf

4. PL Spath and DC Dayton, “Preliminary Screening – Technical and Economic Assessment of Synthesis Gas to Fuels and Chemicals with Emphasis on the Potential for Biomass-Derived Syngas”, http://www.fischer-tropsch.org/DOE/DOE_reports/510/510-34929/510-34929.pdf , NREL/TP-510-34929, 2003.

5. X Li, L Feng, Z Liu, B Zhong, DB Dadyburjor, and DL Kugler, “Higher Alcohols from Synthesis Gas Using Carbon-Supported Doped Molybdenum-Based Catalysts”, Ind. Engr. Chem. Res. 37, 3853-3863, 1998.