Energy from Wastewater

Producing energy from waste is hardly a new idea, as illustrated by efforts such as the landfill methane-to-LNG project I mentioned in November and by numerous demonstration projects to turn various forms of solid waste into liquid fuels. However, I don't normally think of wastewater as having any noteworthy energy potential to extract. An emailed press release from the American Chemical Society announcing the publication of a paper on this subject in Environmental Science & Technology opened my eyes. Based on tests of wastewater streams in Britain, it might be possible to produce at least as much energy from wastewater as is normally used in the process of treating it. That's an intriguing prospect, although we should put it in perspective.

The press release and the paper included a number of statistics, some more accessible than others. One of the two wastewater streams tested in the UK had an energy content of 7.6 kilojoules per liter, which as the ACS noted is equivalent to each gallon being able to light a 100 watt bulb for 5 minutes. (The other sample contained twice that much energy.) That still doesn't sound very impressive until you factor in some other statistics in the report, such as the 12.5 trillion gallons per year of wastewater processed annually in US sewage treatment plants, at an energy cost of around 1.3% of US electricity consumption. That suggests that in theory US treated wastewater might generate as much power, net of processing requirements, as we currently get from biomass power plants fired on wood, which supplied just under 1% of our electricity last year.

Actually harnessing that potential wouldn't be simple. As you might imagine, and as the paper confirms, wastewater is hardly uniform. It contains a wide array of compounds that would require different techniques to produce energy in useful forms such as hydrogen, methane, or direct electricity. The more techniques required at one facility, the bigger the investment involved. Another complication relates to the low concentrations in which these energy precursors are found in wastewater. The team at Newcastle University measured them at levels in the hundreds of parts per million--well below 1%. That implies the need for either separation before processing or handling very large volumes. Either approach adds to cost or complexity. Still, the idea of producing energy from this source at the sewage treatment plant goes one step beyond the usual logic of waste-to-energy projects: Not only is the cost of the feedstock low or effectively negative, but it requires no additional transportation; it's already flowing through the plant.

Extracting energy from wastewater won't solve all of our energy or environmental problems, and it might prove to be more costly than it's worth. However, the paper's conclusion, that this merits a closer examination in a wider variety of settings, seems entirely justified. It makes even more sense in the context of our growing need to make better use of the water portion of this waste.

A Self-Fulfilling Bet on Biofuels?

An article in today's Financial Times (registration required) raises a worrying possibility concerning the plans of the US and other oil-consuming countries to rely on biofuels for an increasing fraction of future fuel needs. What if oil-producing countries took those plans seriously and reduced their investment in new oil capacity, on the assumption that it wouldn't be needed? In some respects, that's exactly what we have in mind. However, if biofuels then failed to materialize in sufficient quantities to fill the gap between oil supply and total fuel demand, or proved to be economically or environmentally unsustainable, then we might inadvertently create precisely the sort of crisis these efforts were intended to avert. It would be easy to dismiss this argument as OPEC-inspired propaganda, if global oil production didn't require enormous ongoing investments to counteract the natural decline rates of producing fields, and if producing-country governments weren't already under internal pressure to spend their oil profits on programs other than reinvesting in future production.

The good news here is that biofuels have reached a scale at which they actually matter in the global oil supply and demand balance. That wasn't the case during the oil crises of the 1970s, and they were still only a marginal factor when oil prices last peaked in 2008. The latest publicly-available issue of the International Energy Agency's Oil Market Report indicates that biofuels now contribute the equivalent of 400,000 barrels per day (bpd) of oil, before including US and Brazilian ethanol volumes that together equate to another 650,000, bringing the global total to just over a million bpd. That might not sound like a large share of a total market of 85 million bpd, but it's enough to influence the global price of oil, which is set at the margin. Doubling or tripling biofuel output would certainly cost oil producers money, if they ignored this factor in their capacity planning.

So far, this is only a problem for oil producers. It becomes a problem for the rest of us when the biofuel plans and targets of consuming countries are based on unproven technology that may not be able to deliver in time, or possibly at all. Unfortunately, that's the position in which we find ourselves. Consider the Renewable Fuel Standard (RFS) enacted by the Congress in 2007 and refined in new regulations issued by the Environmental Protection Agency. Out of the 36 billion gallon per year target for 2022, only around 16 billion gallons is accounted for by corn-based ethanol and first-generation biodiesel--both of which have been amply proven, however much they depend on generous subsidies to remain competitive. 20 billion gallons per year must come from cellulosic ethanol and other advanced biofuels, none of which are in truly commercial production today, in spite of the hype that has been generated by a handful of "demonstration facilities."

One indication of just how unrealistic these targets might be is that EPA was forced to reduce the cellulosic biofuel target it will enforce for 2010 from 100 million gallons to 6.5 million gal.--the equivalent of just over 400 barrels per day of oil--due to lack of supply. And while the agency attributes that shortfall to delays in starting up new facilities using a variety of new technologies, a careful reading of their analysis suggests the problem might be more serious than that. Two firms account for nearly a third of the 694 million gallons of cellulosic biofuel capacity they expect will be in operation by 2014, Cello Energy and Range Fuels. Unfortunately, last year Cello was ordered by a federal court to pay $10 million for defrauding investors concerning its technology claims. Meanwhile blogger Robert Rapier has documented the problems that Range Fuels has experienced in scaling up its process for producing ethanol from gasified biomass. Until both of these firms have demonstrated they can actually do what they claim, at full scale, it's not prudent to bet the ranch on their production forecasts.

Problems such as this are probably just the tip of the iceberg when it comes to scaling up a myriad of new processes for producing motor fuels from non-food biomass, not because it's impossible or because the firms involved don't have sufficient smarts--though one or both of those factors will turn out to apply in at least a few cases--but because it is intrinsically hard. Scientists have been working on cellulosic biofuels and biomass-to-liquids processes for decades, yet the sum total of all that work, up until this point, has only yielded enough fuel production to cover the annual consumption of about 13,000 average American cars. That doesn't mean that companies and investors are foolish to pursue these technologies, or that ExxonMobil is wrong about the potential they apparently see in algae-based fuels, another hot biofuels sector. What it does mean, however, is that when dealing with technologies that can't be made to appear on command and are subject to a number of serious, unresolved technical and logistical challenges, neither consumers nor our governments should base their plans for the future on the assumption they will mostly succeed on schedule.

How realistic is it that the oil-producing countries that control access to the vast majority of the world's oil reserves would be so convinced by our rhetoric concerning biofuels replacing oil, that they will cut back their investments in new capacity? Part of the answer lies in the narrative of Peak Oil that generated headlines when oil prices were spiking a couple of years ago, involving the high decline rates of mature oil fields and the relatively low investment rates of many producing countries. When the government of Venezuela must borrow money from China despite $80 oil, that's one signpost that they might not have enough to reinvest in exploration and production. We can argue about the likely date of a peak in global oil output, but anything that provides governments an excuse to spend less sustaining their oil industries brings that date closer--and that's equally true for a US administration that appears so confident of the success of its biofuels and fuel economy programs that it can allow the timing of the next offshore oil leasing cycle to slip further and further.

Oil is still the lifeblood of our industrial civilization, but it's also a business requiring enormous investments premised on the likelihood of future demand. That doesn't mean we must remain helpless hostages to foreign oil suppliers; fuel efficiency and biofuels are both sensible--even necessary--strategies for us to pursue. But we have an even larger stake in ensuring that the biofuel goals and plans we communicate, not just among ourselves but simultaneously to our oil suppliers, are based on reality. If both we and they are betting on supplies of advanced biofuels that could well fall significantly short of our expectations, then it is we who will suffer the consequences at the gas pump.

The Necessity of "All of the Above"

I've devoted a lot of space in this blog to explaining why we need all of the energy choices currently available to us, including energy efficiency, in order for our economy to have the energy it needs to grow. Although we can't drill our way to energy independence, as some might wish, it's equally clear that the point at which renewable energy sources could enable us to dispense with fossil fuels is still a long ways off. I have no idea how many words I've written in service of this argument, but it's certainly many more than the clichéd 1,000:1 exchange rate vs. the picture that I suddenly realized I've never supplied. Today's posting is a modest attempt to rectify that omission: a simple graph showing how a significant shift in our energy consumption and production patterns might play out between now and 2020.

This picture starts with our total primary energy consumption in 2008 of 99.3 quadrillion BTUs (quads.) Nearly three-fourths of our needs, or 73.7 quads, were produced domestically by a mix of 79% coal, oil and natural gas, 11.5% nuclear power, and a bit over 3% hydropower. Non-hydro renewables--the wind, solar, geothermal and biomass power plus biofuels that constitute the primary focus of US energy policy today--made up the remaining 6.5% of domestic energy production. Now add the 26% of US energy consumption supplied by imports, mainly crude oil and petroleum products, and we have the breakdown shown at the left hand edge of the graph. The rest of the picture is the result of a highly simplified set of assumptions based on phasing out fossil fuels and replacing them with the non-hydro renewables that have been growing so rapidly. It ignores such important considerations as reliability and intermittency, compatibility with infrastructure, and turnover of vehicle fleets.

According to the Energy Information Agency's data, while wind and solar power have been growing at roughly 30% per year each, the total renewables category has been growing at a somewhat slower pace, even after separating out hydropower, which has actually declined significantly since the 1990s. While the average growth rate for all the non-hydro renewables since 2000 has been around 4%, I've more than doubled this for the purposes of my projection to 10%. Renewables would do very well to sustain that kind of pace over the next 11 years, because the bigger they get, the more capital they will require each year to add the next year's increment of growth, and the more hurdles they will face, particularly from NIMBY or "energy sprawl" concerns. 10% compound growth would see these renewables more than triple by 2020, providing plenty of room for biomass/biofuels to double and for wind and solar to double several successive times.

I've also assumed a steady improvement in energy efficiency of 1% per year. If that doesn't sound impressive, compare it against a pre-recession trend of 1% annual growth supporting population growth of around 0.7% and economic growth of 2-3%. 1%/yr. would reduce total energy consumption by almost 12% by 2020, reflecting an improvement in BTUs/$GDP over this interval on the order of around 35%. A recent study by McKinsey & Co. indicated that if the US invested $500 billion in energy efficiency, we could cut our energy consumption by 23% by 2020, so my view is only a little more conservative, reflecting my experience that such things tend to take a bit longer than we expect.

As for nuclear, I think we'll do well to maintain the output of the existing fleet without seeing retirements outweigh additions in this timeframe. Most of the new plants now being discussed would probably only affect the last couple of years of this scenario, in any case.

The biggest impact in the graph above comes from my assumption that domestic fossil fuel production would fall by 5% per year. That will probably seem extreme to some and timid to others. It certainly looks extreme in the context of the recent surge in natural gas production and the stable output of US coal mines. Even US oil production has staged a bit of a comeback recently, thanks to successes in the portions of the Gulf of Mexico where drilling is allowed. However, in the absence of strong sustained rates of oil and gas drilling--drilling that requires both access to resources and a supportive regulatory climate, neither of which appears to be forthcoming--these successes will fade and the high intrinsic decline rates of the mature US hydrocarbon basins would take over. And with new coal-fired power plants being canceled and older ones facing tough competition from gas turbines and renewable power, along with restrictions on practices such as "mountaintop mining" and the prospect of either a Congressionally-mandated or EPA-imposed cap on emissions, a decline in coal output would accompany drops in oil and gas.

All of these growth and decline rates working together produce the picture above, showing fossil fuels tailing off faster than renewables can backfill for some time. That results in net US energy imports growing through 2016, then tapering off gradually as renewables finally gather momentum, but with us arriving at 2020 even less energy independent than we are today. That outcome explains my strong conviction that it is premature for us to give up on the valuable contribution of domestic oil and gas, particularly when we take into account the form that most of those growing energy imports is likely to take: imported oil. Naturally, this is just one scenario among many--though I'd argue it's likelier than some--and it illustrates that no single solution, neither renewables, nor efficiency, nor even greatly expanded drilling, is likely to be capable of delivering the energy we will need without increasing our vulnerability to foreign energy suppliers. They must all work together, if we're to make meaningful progress toward greater energy self-reliance.

The Big Chunks

In my first posting of the year, I suggested that the new administration faced a big decision about our national energy policy, between making either energy security or our response to climate change its top priority. Whichever they choose, President Obama's team needs a framework for implementing that selection, and I'd like to suggest one that appears to go distinctly against the grain of their current thinking. Rather than focusing mainly on renewable energy, a successful US energy strategy must come at this problem from the perspective of its largest elements--the fuels and technologies that account for more than 90% of our current energy supply--not the smallest current pieces, no matter how rapidly they are growing. Renewables are important and destined to become much more so, but we should assess them in terms of their usefulness in displacing those conventional energy sources that are in most urgent need of substitution, as we proceed to reduce emissions or improve energy security.

Oil is still by far our largest energy source, comprised of two main components, domestic production and oil imports. Together they accounted for 39% of the 101.6 quadrillion BTUs (quads) of energy that the US consumed in 2007, split 28%/11% between imports and US production, respectively, though the latter understates the contribution of domestic oil drilling, which also provides around one quarter of our natural gas output. As shown in the pie chart below, natural gas comes next, followed by coal and nuclear power. These five "big chunks" together make up 93% of the total energy we consume, 91% of our electricity generation, and roughly 99% of the energy used for transportation.





Next consider how large renewable energy is likely to grow over the next two decades. The latest forecast from the US Department of Energy anticipates the contribution of wind, solar, geothermal and biomass, including biofuels and biomass-power, to grow from 3.5% of US energy consumption in 2007 to 7.7% by 2030. This estimate still seems fairly conservative, so let's be generous and assume it could reach 15-20%. I know that won't seem aggressive to the most ardent advocates of renewables, but considering the obstacles that would need to be overcome, including making cellulosic biofuel cost-competitively on an industrial scale and expanding non-dispatchable forms of renewable electricity production such as wind and solar into market segments currently dominated by on-demand power from gas turbines and coal plants, it's plenty ambitious. Thus with hard work, dedication and ample investment in the sector, we just might be able to grow renewable energy large enough in 20 years to replace one of those five big blocks I mentioned earlier. Which one to choose depends on the priority of emissions reductions vs. energy security.

The most popular target today would probably be imported oil, widely regarded as the major source of energy insecurity. While obtaining 20% of our future energy from renewables doesn't quite match up with the current 28% from imported oil, with some help from efficiency improvements it just might do the trick. Of course, that would mean that the energy from wind, solar and other electricity-producing technologies would have to displace petroleum products by means of the mass marketing of plug-in electric vehicles. 21 years is not a tremendously long time for that to happen, given the decade that it has taken non-plug-in hybrids to reach a 2.4% market share, but if we're dreaming, let's dream big. So by 2030 we've displaced foreign oil, captured all that money we used to send overseas, and presumably beggared the Middle East. Unfortunately, we have only made a medium-sized dent in our greenhouse gas emissions in the process. Transportation was the number two source of those, and cutting its contribution roughly in half--which is probably optimistic, given the emissions associated with current biofuels--would reduce our 2006 emissions by 15%.

Another popular alternative would be to go after coal's market share. In theory getting 20% of our energy from renewables ought to be sufficient to cut coal use by about 85%--or perhaps only 70% if we allow that we'd be unlikely to burn liquid biofuels such as ethanol and biodiesel to make electricity. Still, cutting 70% of the emissions from coal would reduce US GHG emissions by nearly 25% (very large file), which would put us well on the way to cutting them by 80% by 2050, as envisioned by the administration and Congress. Unfortunately, the fastest-growing renewable energy technologies, wind and solar power, do not compete very well with coal-fired power plants. The latter can run 24/7, providing baseload and mid-load dispatchable power, electricity that is there whenever someone flips the switch, while wind is intermittent, producing power less than a third of the time--often when demand is low--and solar is cyclical and seasonal, making its best contribution into the afternoon demand peak.

When we take these factors into account, wind and solar turn out to be most interchangeable with our least objectionable fossil fuel, natural gas. In this regard, at least, Mr. Pickens' Plan is built on solid ground. But would we really want to displace a large portion of the 23% of our energy we get from natural gas with renewables, or back out the 32% of US natural gas used for electricity, in order to substitute it for petroleum products in cars or elsewhere? At a glance, the energy security and greenhouse gas benefits of this seem less compelling than the direct substitution of renewables for imported oil noted above. And I haven't even mentioned the possibility of using renewables to displace nuclear power, because that would at best net us zero in terms of either energy security or greenhouse gas benefits.

What this framework does, then, is to translate our competing energy priorities into their resultant choices about how best to employ the limited quantities of renewable energy we are likely to have within the next two decades, relative to the major elements of our current energy supply and demand balance. In addition, our preferences for which big chunk of conventional energy on which to focus our efforts will determine which particular renewable energy technologies look most advantageous and deserving of government help in the form of R&D, investment and deployment incentives. From where I sit, this looks like a much more sensible approach than the status quo one that assumes all renewables are equally beneficial in our complex energy economy and that supports them all, indiscriminately.

Ethanol Smoking Gun?

This morning's Washington Post and Wall Street Journal include coverage of important new scientific findings concerning the greenhouse gas benefits of ethanol. They refer to two key papers published in Science on the land-use impact of biofuels. One of these, by Searchinger, Heimlich, et al, suggests that when the global land-use consequences of our diversion of grain into fuel production are considered, the greenhouse gas balance for corn ethanol shifts from a modest reduction to a net doubling, compared to the oil it displaces. And although ethanol from cellulosic sources appears to be less harmful, it may still contribute more to global warming than previously thought. This is a stunning finding, with implications for the entire US energy policy, which prior to the fuel economy and other efficiency provisions of the 2007 Energy Bill has been largely an ethanol policy.

These results are bound to generate controversy, and I would echo calls for them to be confirmed by thorough peer review and additional studies. If they constituted the only concerns about the consequences of conventional ethanol production, prudence would dictate a carefully measured response. However, it has become increasingly clear over the last two years, as US ethanol production and consumption have ramped up dramatically, that it is a mixed blessing, at best. Not only is it no silver bullet for our energy and environmental challenges, but it has also contributed to worrying levels of domestic and global inflation in wholesale and retail food prices.

Public support for ethanol rests on three main benefits: that it enhances energy security, reduces greenhouse gas emissions, and bolsters US agriculture. Of these three attributes, the third is the least ambiguous. Ethanol's energy security contribution turns out to be more a matter of engineering a shift from imported petroleum to the imported natural gas and fertilizer consumed in the cultivation and processing of corn into ethanol, along with a roughly 30% uplift from the solar energy captured by the crops. Its climate change benefits have been viewed as modest but still important, with a generally-accepted figure of a 20% reduction versus gasoline. The latest studies turn that assumption on its head, suggesting that the climate and energy benefits are not actually complementary, but rather trade the former off against the latter. If confirmed by further studies, this conclusion must surely undermine popular and political support for corn-based ethanol.

We haven't had many opportunities to test the relative priority of our concerns about climate change and energy security. The lengthy evolution of the Energy Bill in the Congress last year provided one such occasion, though, when supporters of turning coal into liquid fuels were unable to convince a majority that the energy benefits of this relatively-proven technology outweighed the approximate doubling of greenhouse gas (GHG) emissions it would produce, relative to petroleum products. It's ironic that the GHG outcome for corn ethanol--one of the big winners in the 2007 energy legislation--now looks as bad as coal liquefaction.

In light of the new findings, I believe the most prudent and appropriate course of action would be as follows:

1. Freeze the conventional biofuel portion of the national Renewable Fuel Standard (RFS) in the 2007 Energy Bill at the levels that were in place before its passage. That would re-set the RFS for 2008 to 5.4 billion gallons, rising to 7.5 billion gallons per year in 2012, compared with a requirement for 9 billion gallons this year, rising to 15 billion by 2015.
2. Cap the volume of conventional ethanol eligible for the $0.51/gallon blenders' credit at 7.5 billion gallons per year, effective immediately.
3. Transfer the excess conventional biofuel subsidy--the amount associated with annual conventional biofuel volumes over 7.5 billion gal/yr--to cellulosic ethanol and other advanced biofuels meeting the greenhouse gas savings standards set by the Energy Bill.

Taken together, these actions would provide corn ethanol producers with a modest amount of headroom for further growth, but without a higher mandate to force their output into the market. At the same time, it would offer even stronger support for organizations developing cellulosic ethanol and other advanced biofuels, which still face significant obstacles to commercial production. What we can't do, however, is to pretend that the latest findings--because they are so inconvenient--are somehow irrelevant. They are strong indicators of the need for a major course correction on our path toward energy security that is also compatible with a safer climate. Expect to hear a lot more on this issue in the weeks and months ahead.

Cellulosic Spindletop

Thus far, much of the interest in cellulosic ethanol has focused on the development of enzymes to promote fermentation, rather than on the provision of the crop waste or non-food crops that would be converted into fuel. An article in Monday's San Francisco Chronicle focused on the activity at the "upstream" end of this future value chain. Even if the comparison to a "Spindletop" of cellulose is hyperbole at this point, it's intriguing to consider the risks that these developers are taking on, as they push to develop and plant new energy crops in advance of the full commercialization of the process they are intended to feed.

When I first ran across the research into methods of turning cellulose into ethanol a few years ago, the Department of Energy folks who were promoting it were targeting "corn stover", the waste from corn ethanol production, as their likely initial feedstock. That still probably makes sense, since this material will already have been harvested and collected in the course of producing the feed for current-generation ethanol plants. It's not hard to imagine first and second generation facilities operating side by side, until the cellulosic process becomes cheap enough to displace the grain-based version--if ever. However, others are looking beyond corn stover to switch grass and the miscanthus hybrid mentioned in the Chronicle. The allure of this tall, rapidly growing "superweed" is understandable, although those pioneering this energy crop will have to manage two key issues.

First, they must find alternative outlets for their production, in case of delays in making cellulosic ethanol processing fully operational. That probably means lining up biopower plants to consume their biomass. This sector, which involves firing or co-firing smallish thermal power plants with wood and other biomass, has been a modest success, despite receiving much less publicity than wind or solar power. The key hurdle involved would be proximity, given the economics of transporting low-density fuels long distances.

The other issue concerns real or perceived hazard to other crops. Even if the ultimate variety of miscanthus chosen for energy cropping doesn't involve genetic modification, there's a long history in this country of foreign plants that have gotten out of control or failed to work out as intended. In California brittle, messy Eucalyptus trees were introduced from Australia as wind breaks. I'm particularly attuned to this example, since one of these ill-considered trees dropped a 9-ton limb on our house when I was growing up. The southern experience of kudzu is probably even more relevant to miscanthus.

Whatever the ultimate feedstock limit on biofuel production, it's clear that that limit will be much higher if it is based on a wide variety of hardy, low-input energy crops, rather than on a few food crops. Once the cellulosic ethanol process has been fully demonstrated at an industrial scale and the cost of its enzymes falls sufficiently to make it competitive with corn ethanol, its ultimate success will depend on the availability of high-quality, low-cost biomass. Outside the corn belt, that will mean dedicated energy crops such as miscanthus and switchgrass, if their commercial risks can be managed successfully. Perhaps the players in the different segments of this nascent value chain should start considering vertical integration or alliances.