Energy Generation Now & in the Future

I finished my last post with an important piece of data.  Out of 100 quads of energy the US generates every year, the vast majority of it (83%) comes from fossil fuel sources – sources that emit greenhouse gases when we burn them.  The same is true for the vast majority of other countries, and therefore for the global portfolio as well.  Here is a graphic showing global energy consumption distribution by fuel type from 1990 through 2010 and into the future:

Figure 1. Global fuel-type energy consumption, 1990-2040 (EIA 2013 Energy Outlook).

The global picture is somewhat different from the US picture: liquids’ energy (e.g., oil) exceed coal energy, which exceed natural gas.  All three of these carbon-intensive energy sources, which power our developed, high-wealth lifestyles, greatly exceed renewables (which hydropower dominates), which exceeds nuclear.  It is these type of energy forecasts that lead to the suite of IPCC emissions pathways:

Figure 2. IPCC Fifth Assessment Report Representative Concentration Pathway (RCP) CO2-eq concentrations.

Note that our current emissions trajectory more closely resembles the RCP8.5 pathway (red) than the other pathways.  This trajectory could lead to a 1000+ ppm CO2-eq concentration by 2100, or 2.5X today’s concentration value.  Stabilizing global temperature increases at less than 2C by 2100 requires stabilizing CO2-eq concentrations below 450 and quickly decreasing, which is best represented by the RCP2.6 pathway above (green).  This pathway is technologically impossible to achieve as of today.  The only way to make it possible is to invest in innovation: research, development, and global deployment of low-carbon technologies.  We are not currently doing that investment; nor does it look likely we will in the near future.

Let’s take a further look at the recent past before we delve further into the future.  Environmental and renewable energy advocacy groups tout recent gains in renewable energy deployment.  We should quietly cheer such gains because they are real.  But they are also miniscule – far too little deployment at a time when we need exclusive and much wider deployment of renewable energy globally to shift our emissions pathway from RCP8.5 to RCP2.6.  Here is a graphic showing global use of coal in the past 10+ years:

Figure 3. Global coal use in million tonnes of oil-equivalent 2001-2011 (Grist).

Climate and clean energy advocates like to report their gains in percentage terms.  This is one way of looking at the data, but it’s not the only way.  For instance, coal usage increased by 56% from 2001 to 2011.  This is a smaller percentage than most renewable energy percentage gains in the same time period, but the context of those percentages is important.  As you’ll see below, renewable energy gains really aren’t gains in the global portfolio.  The above graph is another way to see this: if renewable energy gains were large enough, they would replace coal and other fossil fuels.  That’s the whole point of renewable energy and stabilizing carbon emissions, right?  If there is more renewable energy usage but also more coal usage, we won’t stabilize emissions.  Here is another way of looking at this statement:

Figure 4. Global Energy Consumption from Carbon-Free Sources 1965-2012 (Breakthrough).

Carbon-free energy as a part of the total global energy portfolio increased from 6% in 1965 to 13% in the late 1990s.  This is an increase of 200% – which is impressive.  What happened since the 1990s though?  The proportion was actually smaller in 2011 than it was in 1995 in absolute terms.  At best, carbon-free energy proportions stagnated since the 1990s.  Countries deployed more carbon-free energy in that time period, but not enough to increase their proportion because so much new carbon energy was also deployed.  What happened starting in the 1990s?  The rapid industrialization of China and India, predominantly.  Are developing countries going to stop industrializing?  Absolutely not, as Figure 1 showed.  It showed that while renewable energy consumption will increase in the next 30 years, it will likely do so at the same rate that natural gas and liquids will.  The EIA projects that the rate of increase of coal energy consumption might level off in 30 years, after we release many additional gigatonnes of CO2 into the atmosphere, ensuring that we do no stabilize at 450 ppm or 2°C.

Here is the EIA’s projection for China’s and India’s energy consumption in quads, compared to the US through 2040:

Figure 5. US, Chinese, and Indian energy consumption (quads) 1990-2040 (EIA 2013 Energy Outlook).

You can see the US’s projected energy consumption remains near 100 quads through 2040.  China’s consumption exceeded the US’s in 2009 and will hit 200 quads (2 US’s!) by 2030 before potentially leveling off near 220 quads by 2040.  India’s consumption was 1/4 the US’s in 2020 (25 quads), and will likely double by 2040.  Where will an additional 1.5 US’s worth of energy come from in the next 30 years?  Figure 1 gave us this answer: mostly fossil fuels.  If that’s true, there is no feasible way to stabilize CO2 concentrations at 450 ppm or global mean temperatures at 2°C.  That’s not just my opinion; take a look at a set of projections for yourself.

Here is one look at the future energy source by type:

Figure 6. Historical and Future Energy Source by Type (BNEF).

This projection looks rosy doesn’t it?  Within 10 years, most new energy will come from wind, followed by solar thermal.  But look at the fossil fuels!  They’re on the way out.  The potential for reduced additional fossil fuel generation is good news.  My contention is that it isn’t happening fast enough.  Instead of just new energy, let’s look at the cumulative energy portfolio picture:

Figure 7. Historical and Future Total Energy Source by Type (BNEF).

This allows us to see how much renewable energy penetration is possible through 2030.  The answer: not a lot, and certainly not enough.  2,000 GW of coal (>20% of total) remains likely by 2030 – the same time when energy experts say that fossil fuel use must be zero if CO2 concentrations are to remain below 450 ppm by 2100.  But coal isn’t the only fossil fuel and the addition of gas (another 1,700 GW) and oil (another 300 GW) demonstrates just how massive the problem we face really is.  By 2030, fossil fuels as a percentage of the total energy portfolio may no longer increase.  The problem is the percentages need to decrease rapidly towards zero.  Nowhere on this graph, or the next one, is this evident.  The second, and probably more important thing, about this graph to note is this: total energy increases at an increasing rate through 2030 as developing countries … develop.

Figure 8. Global fuel-type energy consumption, 1990-2040 (EIA 2013 Energy Outlook).

The EIA analysis agrees with the BNEF analysis: renewables increase through 2030.  The EIA’s projection extends through 2040 where the message is the same: renewables increase, but so do fossil fuels.  The only fossil fuel that might stop increasing is the most carbon intensive – coal – and that is of course a good thing.  But look at the absolute magnitudes: there could be twice as many coal quads in 2040 as there were in 2000 (50% more than 2010).  There could also be 50% more natural gas and 30% more liquid fuels.  But the message remains: usage of fossil fuels will likely not decline in the next 30 years.  What does that mean for CO2 emissions?

Figure 9. Historical and projected global carbon dioxide emissions: 1990-2040 (EIA 2013 Energy Outlook).

Instead of 14 Gt/year (14 billion tonnes per year) in 2010, coal in 2040 will emit 25 Gt/year – almost a doubling.  CO2 emissions from natural gas and liquids will also increase – leading to a total of 45 GT/year instead of 30 GT/year.  The International Energy Agency (IEA) estimated in 2011 that “if the world is to escape the most damaging effects of global warming, annual energy-related emissions should be no more than 32Gt by 2020.”  The IEA 2012 World Energy Outlook Report found that annual carbon dioxide emissions from fossil fuels rose 1.4 percent in 2012 to 31.6 Gt.  While that was the lowest yearly increase in four years, another similar rise pushes annual emissions over 32Gt in 2014 – six years ahead of the IEA’s estimate.  Based on the similarity between our historical emissions pathway and the high-end of the IPCC’s AR4 SRES scenarios (see figure below), 2°C is no longer a viable stabilization target.

Figure 10. IEA historical annual CO2 emissions and IPCC AR4 emissions scenarios: 1990-2012 (Skeptical Science).

The A2 pathway leads to 3 to 4°C warming by 2100.   Additional warming would occur after that, but most climate science focus ends at the end of this century.  A huge caveat applies here: that warming projection comes from models that did not represent crysophere or other processes.  This is important because the climate system is highly nonlinear.  Small changes in input can induce drastically different results.  A simple example of this is a change in input from 1 to 2 doesn’t mean a change in output from 1 to 2.  The output could change to 3 or 50, and we don’t know when the more drastic case will take place.  Given our best current but limited understanding of the climate system, 3 to 4°C warming by 2100 (via pathway A2) could occur.  Less warming, given the projected emissions above, is much, much less likely than more warming than this estimate.  Policy makers need to shift focus away from 2°C warming and start figuring out what a 3 to 4°C warmer world means for their area of responsibility.  Things like the timing of different sea level rise thresholds and how much infrastructure should we abandon to the ocean?  Things like extensive, high-magnitude drought and dwindling fresh water supplies.  These impacts will have an impact on our lifestyle.  It is up to us to decide how much.  The graphs above and stories I linked to draw this picture for me: we need to change how we approach climate and energy policy.  The strategies employed historically were obviously inadequate to decarbonize at a sufficient rate.  We need to design, implement, and evaluate new strategies.

Climate & Energy Articles – Aug. 17, 2013

Since I can’t devote as much time to everything I read, here is a quick roundup of things I thought were interesting recently:

A Nature article (subs. req’d) describes some of the problems with a trending climate effort: decadal predictions.  In the past, agencies just made climate projections for a couple of centuries into the future.  In addition to that, interest in projections over the next 10 or 20 years grew.  Unfortunately, climate models aren’t well designed for these short time frames.  Thus, they miss high-frequency climate events made just after agencies issue them.  Of particular concern are the high impact events, as we tend to focus on them.  I would remind critics that point out these “misses” that very few financial models indicated the biggest economic disruption of our lifetime: the Great Recession, yet we continue to ascribe great status to the same financial titans that universally missed that high impact event.  That means, of course, that critics remain within their tribal identities and look for any evidence to support their position, even as they ignore similar evidence for analogous cases.

A group made an interesting counter argument regarding the cause behind the US’s recent drop in CO2 emissions.  Instead of the switch from carbon-intensive coal to slightly less intensive natural gas, as many analysts described, this group claims the drop occurred due to widespread, massive efficiency gains.  I characterize this as interesting because the group is countering the International Energy Agency, among others.  While not prescient, the IEA is the leading authority in these types of analyses.  We shouldn’t take their analyses without a grain of salt, of course, as their methodologies are likely imperfect.  Instead, this new argument should encourage further research and analysis.  Was the coal-to-gas switch primarily responsible or was efficiency?  Additional years’ data will help to clarify the respective roles.  In the long-term, efficiency can play as big or a bigger role than the coal-to-gas switch that occurred to date.  That’s where innovation funded from a carbon price comes into play.

Grist ran an informative series recently that included a short video of how much energy the US uses – the primary generators and consumers by type and sector.  The upshot is this: the US uses 100 quads (an energy measurement), which makes further discussion quite simple.  The US generates 81-83 quads (81-83%) via fossil fuels (oil, coal, and natural gas).  That leaves only 17-19% of US generation by non-fossil sources.  Most non-fossil energy generation is nuclear, which means renewables account for the smallest share of energy generation.  Most of that is hydropower from dams that we built in the first half of the 20th century.  This data will form the basis of my next post, which will examine the implications of this energy breakdown for climate policy.  What will it take to replace 83 quads of fossil fuel energy generation with renewable energy generation?

45.3% of the Contiguous United States in Moderate or Worse Drought – 15 Aug 2013

According to the Drought Monitor, drought conditions improved recently across some of the US. As of Aug 15, 2013, 45.3% of the contiguous US is experiencing moderate or worse drought (D1-D4), as the early 2010s drought continues month after month.  This value is about 10 percentage points lower than it was in the early spring. The percentage area experiencing extreme to exceptional drought increased from 14.6% to 14.8%; this is approximately 4% lower than it was six months ago. The eastern third of the US was wetter than normal during July into August, which helped keep drought at bay.  The east coast in particular was much wetter than normal and the summer monsoon was much more active this summer compared to 2012.  Instead of Exceptional drought in Georgia and Extreme drought in Florida two years ago, there is flash flooding and rare dam water releases in the southeast.  Four eastern states experienced their top-four wettest Julys on record.  The West presents a different story.  Long-term drought continues to exert its hold over the region, as it remained warmer than normal but six southwestern states received top-20 July precipitation this year.  Meanwhile, Oregon recorded its driest July on record.

Figure 1 – US Drought Monitor map of drought conditions as of August 13th.

If we compare this week’s maps with previous dates (here and here, for example), we can see recent shifts in drought categories.  Compared to early July, and despite recent rain events, drought expanded in the Midwest (into Iowa, Missouri, Illinois, and Minnesota) as well as Louisiana, Arkansas, and Mississippi.

Here is the Western US drought map this week:

Figure 2 – US Drought Monitor map of drought conditions in Western US as of August 15th.

After worsening during late winter into spring 2013, drought conditions steadied during the past month.  The differences between this map and early July’s is the spatial shift of conditions; the total percent area values are about the same.

Temporary drought relief occurred over parts of Arizona and Colorado as the summer monsoon brought moisture northward and interacted with cooler air masses than normal from Canada.

Here are the current conditions for Colorado:

Figure 3 – US Drought Monitor map of drought conditions in Colorado as of July 9th.

There is clear evidence of relief evident over the past three months here.  Severe drought area dropped from 72% to 69% (this was 100% about six months ago!).  Extreme drought area dropped slightly from 27% to 26% (also down from 50%+ six months ago).  Exceptional drought is down significantly from three and six months ago.  Instead of 17% of Colorado, Exceptional drought now covers only 3% of the state.  The good news for southeastern Colorado was the recent delivery of substantial precipitation.  I didn’t think it would be enough to alleviate the worst conditions, but they received enough precipitation that drought conditions improved from Exceptional to Extreme.  Their drought is not over yet, but they are finally trending in a good direction.  And for the first time in over one year, some small percentage (1%) of Colorado does not currently have any drought condition.  This is great news – hopefully this area expands throughout the rest of the year.

US drought conditions are more influenced by Pacific and Atlantic sea surface temperature conditions than the global warming observed to date.  Different natural oscillation phases preferentially condition environments for drought.  Droughts in the West tend to occur during the cool phases of the Interdecadal Pacific Oscillation and the El Niño-Southern Oscillation, for instance.  Beyond that, drought controls remain a significant unknown.  Population growth in the West in the 21st century means scientists and policymakers need to better understand what conditions are likeliest to generate multidecadal droughts, as have occurred in the past.  Without comprehensive planning, millions of people dwindling fresh water supplies will threaten millions of people.  That very circumstance is already occurring in western Texas where town wells are going dry.  An important factor in those cases is energy companies’ use of well water for natural gas drilling.  This presents a dilemma more of us will face in the future: do we want cheap energy or cheap water?  In the 21st century, both options will not be available at the same time as they were in the 20th century.  This presents a radical departure from the past.

As drought affects regions differentially, our policy responses vary.  A growing number of water utilities recognize the need for a proactive mindset with respect to drought impacts.  The last thing they want is their reliability to suffer.  Americans are privileged in that clean, fresh water flows every time they turn on their tap.  Crops continue to show up at their local stores despite terrible conditions in many areas of their own nation (albeit at a higher price, as found this year).  Power utilities continue to provide hydroelectric-generated energy.

That last point will change in a warming and drying future.  Regulations that limit the temperature of water discharged by power plants exist.  Generally warmer climate conditions include warmer river and lake water today than what existed 30 years ago.  Warmer water going into a plant either means warmer water out or a longer time spent in the plant, which reduces the amount of energy the plant can produce.  Alternatively, we can continue to generate the same amount of power if we are willing to sacrifice ecosystems which depend on a very narrow range of water temperatures.  As with other facets of climate change, technological innovation can help increase plant efficiency.  I think innovation remains our best hope to minimize the number and magnitude of climate change impacts on human and ecological systems.

Research: Timing Sea-Level Rise Projections

A recent surge (see my last post) of published papers on sea-level rise generated the following paper: “Ice sheet collapse following a prolonged period of stable sea level during the last interglacial” (subs. req’d) by O’Leary et al. in Nature Geoscience.  The authors embarked on this work to provide information on potential events in our future as sea levels rise in response to increasing greenhouse gas concentrations.  From their abstract: “During the last interglacial period, 127–116 kyr ago, global mean sea level reached a peak of 5–9  m above present-day sea level.”  An “interglacial” is a time period in between periods with extensive polar ice sheets.  The last interglacial is also known as the Eemian.  During the Eemian, sea levels were 16-30 feet higher than today’s level.  Scientists estimate that 4 feet of additional sea level rise is already locked-in due to past GHG emissions.  Our current emissions pathway through 2100 locks in 23 feet of future sea level rise (over the course of the next few hundred years).

Most analysis of future sea-level rise suggests that the rise is likely to occur over the course of decades to centuries.  The results are dramatic – more than 1,400 coastal US municipalities would be mostly underwater with an additional 23 feet of sea level – regardless of the timescale in which the seas rise.  Society can respond more easily if the time is century-scale than if it is decadal-scale.  The costs are high regardless, but societal strife might be less if the timescale is longer.  Imagine telling one million people they have to abandon their homes in the next decade due to rising seas while their governments are forced to move infrastructure such as roads and airports (billions of dollars’ worth) and businesses have to move inland as well.  I don’t think the result would be pretty.

O’Leary et al. focused on an even more striking event within the Eemian (which was as warm as projections of likely late 21st century temperatures): a rapid, not gradual, increase in sea levels.  “We show that between 127 and 119 kyr ago, eustatic sea level remained relatively stable at about 3–4 m above present sea level. […] We suggest that in the last few thousand years of the interglacial, a critical ice sheet stability threshold was crossed, resulting in the catastrophic collapse of polar ice sheets and substantial sea-level rise.”

Their results (along with others’) show that sea levels were stable a few meters above the current level.  Then, something caused polar ice sheets to melt “rapidly” and sea levels to rise quickly from 3-4 m above today’s to 9 m above today’s. So sea levels rose from 10-12 feet above today’s to nearly 30 feet above today’s in less than one thousand years.  If such an event happens again in our future, one thousand years is a relatively long time to respond to the primary change.  Secondary changes might challenge our responses, but that isn’t my focus today.

If O’Leary’s estimate of one thousand years is too long by one order of magnitude, serious challenges would exist.  If sea levels rise by 17 feet in less than one hundred years, how would we respond?  How much infrastructure would we leave to the rising seas because moving it is just too costly?  In addition to continued sampling of former underwater sites in Australia and elsewhere to gauge total sea level change, scientists need to refine methods to pin down the timing of past changes.  How frequently did past rapid changes occur?  What were their magnitude?  Then, decision makers need to establish two-tiered response plans.  The first should address the most common type of change: gradual rise/fall over hundreds to thousands of years.  The second therefore should address the more serious challenge of sub-centennial rapid sea level change.  We can’t implement plans that aren’t formulated beforehand.

It is worth noting that Mercer first wrote about polar ice sheet collapse 35 years ago.  This phenomenon is not new but is being better informed by subsequent research.  Hat tip New York Times.

July 2013 CO2 Concentrations: 397.23 ppm

During July 2013, the Scripps Institution of Oceanography measured an average of 397.23 ppm CO2 concentration at their Mauna Loa, Hawai’i Observatory.

This value is important because 397.23 ppm is the largest CO2 concentration value for any July in recorded history.  This year’s July value is 2.90 ppm higher than July 2012′s!  Month-to-month differences typically range between 1 and 2 ppm.  This year-to-year jump is clearly well outside of that range.  This change is in line with other months this year: February’s year-over-year change was +3.37 ppm and May’s change was +3.02 ppm.  Of course, the unending trend toward higher concentrations with time, no matter the month or specific year-over-year value, as seen in the graphs below, is more significant.

The yearly maximum monthly value normally occurs during May. This year was no different: the 399.89ppm concentration in May 2013 was the highest value reported this year and, prior to the last five months, in recorded history (neglecting proxy data).  I expected May of this year to produce another all-time record value and it clearly did that.  May 2013′s value will hold onto first place all-time until February 2014.

Figure 1 – Time series of CO2 concentrations measured at Scripp’s Mauna Loa Observatory in July from 1958 through 2013.

CO2Now.org added the `350s` and `400s` to the past few month’s graphics.  I suppose they’re meant to imply concentrations shattered 350 ppm back in the 1980s and are pushing up against 400 ppm now in the 2010s.  I’m not sure that they add much value to this graph, but perhaps they make an impact on most people’s perception of milestones within the trend.

How do concentration measurements change in calendar years?  The following two graphs demonstrate this.

Figure 2 – Monthly CO2 concentration values (red) from 2009 through 2013 (NOAA). Monthly CO2 concentration values with seasonal cycle removed (black). Note the yearly minimum observation occurred nine months ago and the yearly maximum value occurred two months ago. CO2 concentrations will decrease through October 2013, as they do every year after May.

Figure 3 – 50 year time series of CO2 concentrations at Mauna Loa Observatory (NOAA).  The red curve represents the seasonal cycle based on monthly average values.  The black curve represents the data with the seasonal cycle removed to show the long-term trend.  This graph shows the recent and ongoing increase in CO2 concentrations.  Remember that as a greenhouse gas, CO2 increases the radiative forcing of the Earth, which increases the amount of energy in our climate system.

CO2 concentrations are increasing at an increasing rate – not a good trend with respect to minimizing future warming.  Natural systems are not equipped to remove CO2 emissions quickly from the atmosphere.  Indeed, natural systems will take tens of thousands of years to remove the CO2 we emitted in the course of a couple short centuries.  Human systems do not yet exist that remove CO2 from any medium (air or water).  They are not likely to exist for some time.  So NOAA will extend the right side of the above graphs for years and decades to come.

This month, I will return to some graphs I’ve presented before.  Here is a 10,000 year view of CO2 concentrations from ice cores to compare to the recent Mauna Loa observations:

Figure 4 – Historical CO2 concentrations from ice core proxies (blue and green curves) and direct observations made at Mauna Loa, Hawai’i (red curve).

Or we could take a really, really long view:

Figure 5 – Historical record of CO2 concentrations from ice core proxy data, 2008 observed CO2 concentration value, and 2 potential future concentration values resulting from lower and higher emissions scenarios used in the IPCC’s AR4.

Note that this graph includes values from the past 800,000 years, 2008 observed values (8-10ppm less than this year’s average value will be) as well as the projected concentrations for 2100 derived from a lower emissions and higher emissions scenarios used by the 2007 IPCC report.  If our current emissions rate continues unabated, it looks like a tripling of average pre-industrial concentrations will be our future reality (278 *3 = 834).  This graph also clearly demonstrates how anomalous today’s CO2 concentration values are.  It further shows how significant projected emission pathways are when we compare them to the past 800,000 years.  It is important to realize that we are currently on the higher emissions pathway (towards 800+ppm).

The rise in CO2 concentrations will slow down, stop, and reverse when we decide it will.  It depends primarily on the rate at which we emit CO2 into the atmosphere.  We can choose 400 ppm or 450 ppm or almost any other target (realistically, 350 ppm seems out of reach within the next couple hundred years).  That choice is dependent on the type of policies we decide to implement.  It is our current policy to burn fossil fuels because we think doing so is cheap, although current practices are massively inefficient and done without proper market signals.  We will widely deploy clean sources of energy when they are cheap; we control that timing.  We will remove CO2 from the atmosphere if we have cheap and effective technologies and mechanisms to do so, which we also control to some degree.  These future trends depend on today’s innovation and investment in research, development, and deployment.  Today’s carbon markets are not the correct mechanism, as they are aptly demonstrating.  But the bottom line remains: We will limit future warming and climate effects when we choose to do so.