This story is included with an NYT Opinion subscription. Learn more »
Steven Davis of the University of California, Irvine, and Robert Socolow of Princeton (best known for his work dividing the climate challenge into carbon “wedges”) have written “Commitment accounting of CO2 emissions,” a valuable new paper in Environmental Research Letters showing the value of shifting from tracking annual emissions of carbon dioxide from power plants to weighing the full amount of carbon dioxide that such plants, burning coal or gas, could emit during their time in service.
This makes sense because of the long lifetime of these plants once built — typically 40 years or so — and the long lifetime of carbon dioxide once released. (I’d love to see some data visualization experiments on this idea from Adam Nieman, building on his work showing the volume of daily CO2 emissions from cities and the like.)
Here’s Davis’s “video abstract” (the transcript is appended at the end of this post, along with a rich discussion of the paper’s findings and implications):
The opening section of the paper is remarkably clear and is worth posting here (minus footnotes and the like):
Each year, governments and firms estimate and report CO2 emissions from the burning of fossil fuels, and their efforts to slow climate change are measured against these annual emissions. Ultimately, though, the magnitude of warming we experience will not be determined by emissions in any one year, but by cumulative CO2 emissions. Thus, climate scientists and energy-economic modelers have developed hundreds of plausible scenarios of future emissions and used them to identify emissions pathways that might achieve climate policy goals. Such scenarios are powerful tools for connecting emissions and warming to trajectories of population growth, economic development, and energy use. However, these trajectories are constrained by tremendous socio-economic inertia (e.g., existing energy, transport and built infrastructures) that limits the rate at which CO2 emissions can be reduced and climate change avoided.
In 2010, Davis et al [background] quantified an important component of socio-economic inertia by estimating the future emissions expected from all existing fossil fuel-burning infrastructure worldwide, naming these ‘committed’ emissions. That paper provided a single data point (commitments as of 2009), but lacked the context perhaps most important to policymakers: how these committed emissions have changed over time. Here we provide that context. We show that, despite international efforts to reduce CO2 emissions, total remaining commitments in the global power sector have not declined in a single year since 1950 and are in fact growing rapidly (by an average of 4% per year 2000–2012).
The annual ‘commitment accounting’ that we demonstrate here offers policymakers an opportunity to evaluate historical trends and to quantify the long-term consequences of current actions in a new way.
Environmental Research Web has posted “Why our carbon-dioxide emissions are like credit-card debt,” which explains how it grew out of earlier work by Davis and others (Davis, Caldeira and Matthews, Science, 2010). This quote from Davis is particularly helpful:
“One of the things that makes climate change such a difficult problem is that it lacks immediacy,” Steven Davis of the University of California, Irvine, told environmentalresearchweb. “It’s going to have huge impacts in the long run, but its effects on our day-to-day lives seem small. The way we’ve been tracking carbon-dioxide emissions reinforces this remoteness: the annual emissions we monitor are small relative to the cumulative emissions that will cause large temperature increases. The alternative we present, what we call commitment accounting, helps by quantifying the long-run emissions related to investment decisions made today.”
I sought reactions from a host of energy and climate analysts, kicking things off with this query and thought (some email shorthand is cleaned up):
You are probably aware of “Commitment Accounting of CO2 Emissions,” a valuable new Socolow/Davis paper (building on Davis, Caldeira, Matthews, 2010).
Given the near-term and enduring benefits of electric power expansion in developing countries, the other long-term effect of expanded coal-powered generation is accrued wealth and economic growth (along with health costs if they are dirty plants, of course). You could call it a gigawatt-hours commitment.
It’d be interesting to visualize all of this side by side with the emissions commitment in some way.
I think this emerging form of emissions accounting provides a valuable way to show how the growing coal (and natural gas) greenhouse-gas emissions commitment will play out, but — because of the competing social and economic values embedded in that extracted energy, along with the equity argument poor countries use against established fossil-powered industrial giants — I’m not sure it leads to a more effective strategy for cutting those emissions.
I’d love to include your thoughts on this, including links to relevant background that bolsters your points…. On complex issues the best way to pinpoint “reality” is through discourse. As in celestial navigation, the more “lines of position,” the tighter the resulting area on the chart.
(Please excuse the acronyms below. For now, please use Google to find definitions. I don’t have time at the moment to include explanatory links.)
Christopher Green, an energy-focused economist at McGill, was first to reply:
Without taking anything away from the importance of the Davis-Socolow contribution, your suggestion that we should also look at the benefits of the electricity generated is, in my opinion, dead on. The world as a whole (especially the emerging/developing country component) is going to require increased energy consumption for the foreseeable (and likely more distant) future. That that energy, however produced, provides enormous benefits cannot be denied
The important question is how the world’s huge and growing energy requirements are going to be met. Hoffert et al (1998) [the paper is here] provided what still is the clearest framework for establishing the huge magnitude of the energy technology challenge to meeting a growing energy commitment—a challenge measured in terawatts not gigawatts. That challenge has been largely ignored, with the policy focus placed on emissions and emissions reduction and the political will to reduce them, without due regard to the current limits on alternative low carbon energy technologies. The failure to address the Hoffert et al energy technology question is reflected in the fact that since 2000 (and including 2013) the share of global energy consumption accounted for by fossil fuels has remained essentially constant at 86.5-87.0%. (Here I use BP Energy Statistics.) globally, the “progress” made with non-hydro renewables is offset by a decline in energy from nuclear power plants. Not surprisingly, the carbon content of energy is the same in 2013 as it was in the early 1990s and actually rose a little since 2000.
While Davis and Socolow are clearly right that the climate has not benefited from the lack of progress on the energy technology front, billions of people, particularly in the emerging country/developing world, have benefited substantially from the energy generated. As far as I can see the standoff will continue until there is recognition that climate change is first and foremost an energy technology problem, one that cannot be solved/resolved without a great deal more than emission reduction pledges.
David Hawkins of the Natural Resources Defense Council was concerned that I was using too wide a brush:
You say— “Given the near-term and enduring benefits of electric power expansion in developing countries, the other long-term effect of expanded coal-powered generation is accrued wealth and economic growth…”
This statement conflates three things: electrification services, electric generating capacity, and coal-fired generating capacity. The implication is that there is a serious tradeoff between constraining cumulative global CO2 emissions and meeting the needs of developing countries for increased electrification services. Truth is, the CO2 commitment from new generating capacity in the poorest countries will be small, even if they build mostly coal. But building mostly coal is not the only path they or other countries need to follow to expand electrification services. When the world’s wealthiest countries begin to take climate protection seriously, there are ways in which any incremental costs of pursuing a low-carbon electrification path in poorer countries can be shared based on all countries’ strategic interests in avoiding a disrupted climate.
Steve Davis, one of the paper authors, added this:
Dave is right to point out that the largest commitments we’ve quantified are only distantly related to meeting the energy needs of the underserved. The big commitments reflect coal-based industrialization and energy-intensive development whose value is more questionable given the climate and health impacts entailed and availability of cleaner and more energy-efficient options.
I responded with this question:
But given urbanization trends, particularly, centralized power production and the jobs and output that come with industrialization are certainly a big factor for many governments, right?
Davis replied:
Perhaps, but that’s the distant relation. We point out that the world built 89 GW [gigawatts, or billion watts] of coal-fired capacity per year 2010-2012. That’s roughly equal to the entire generating capacity of sub-Saharan Africa being added every year. These plants are supporting large-scale industrialization that is a world apart from providing basic energy services, even if one dreams of the other. They’re related just as the merits of different graduate schools are in the minds of parent teaching their toddlers to read.
Chris Green responded to Davis and Hawkins:
Yes, Dave is talking about the underserved in the sense of the poorest many of whom are not even hooked up to the grid. But my comment addresses the huge energy demands as people move beyond the very poorest. Think here of the many countries, especially the populous ones, that are industrializing and urbanizing. They inevitably will require a lot of energy, if for no other reason than that materials for construction (steel, cement, flat glass, aluminum, copper) are hugely energy intensive — with energy intensities an order of magnitude higher than that for most other manufactured goods. Think too about the demand for appliances including air conditioning as people move into the middle class. These are key factors in the rise in energy demand and the observed increase in fossil fuel generating capacity.
I also take issue with Dave’s suggestion of good alternatives to fossil fuel generated energy. With the possible exception of nuclear energy, which to date has faced numerous hurdles, there certainly aren’t on anything like the scale required. And little has changed in this respect since Hoffert et al, 1998.
Alex Trembath at the Breakthrough Institute joined the general discussion:
Thanks for the note. As you know, Breakthrough and our colleagues at the Consortium for Science, Policy & Outcomes (and development/energy experts from across the world) addressed this issue in our April report “Our High-Energy Planet.” As Professor Green mentioned, the singular focus on emissions is pervasive, and in some cases understandable (for instance, comparing national emissions accounts — measured in MTCO2 — is often much easier than harmonizing and comparing trade-adjusted energy consumption — measured variously in Mtoe, GWhs, bbls, EJs, etc etc). But in the case of energy in emerging economies, the laser focus on emissions is inappropriate.
There has been some movement towards a “gigawatt-hours commitment,” as you write. This is the Decade of Sustainable Energy for All, after all. Unfortunately, the dominant frameworks used for understanding the energy access challenge — including those of the UN and IEA — are pretty clearly unacceptable. There are two main problems. The first is an almost exclusive focus on basic household electricity access, often in remote areas, without attendant attention to urbanizing populations, industrializing economies, etc. The second is the scale of ambition. As most on this thread will know, energy access is defined at 250-500kWh/year. This compares to the average German consumption of over 7000kWh/yr. With German levels of consumption, a planet of 9-10 billion in midcentury would require three times as much energy as the world does today.
If we don’t start with this very basic, fundamental point, then we’re not really having a discussion about energy development or climate change. Again, as Professor Green writes, these realities are commonly ignored, even though folks like Marty Hoffert, Nathan Lewis, Richard Smalley, and John Holdren pointed them out over a decade ago.
Morgan Bazilian and Roger Pielke, Jr. have described what it actually take for sub-Saharan Africa to reach levels of energy consumption enjoyed by South Africans or Americans. China, India, Indonesia, South Africa, and other emerging countries have of course developed largely with coal-fired power. Africa appears poised to mostly leapfrog coal straight towards large hydro and natural gas. No one has yet conceived of a feasible energy development strategy — to power cities, industries, major infrastructure, etc — with renewables or nuclear as the dominant energy sources. Fortunately, a lot of the much-needed work on delivering truly scalable and affordable zero-carbon alternative is being done in the developing world itself. China is pursuing nuclear, solar, and carbon capture at a much greater scale than most Western economies. There are several dozen other countries investing in next-generation nuclear technologies (along with things like renewables, shale gas, coal-to-gas, etc.) to power rapidly growing demand. If decarbonization is to remain a priority — as it should — the energy innovation momentum of the developing world must be captured and accelerated.
I would finally recommend the work of Catherine Wolfram at UC Berkeley, who has both criticized the energy projections methods of the IEA and most recently introduced the very helpful framework of “under-grid” populations, which are very near but unconnected to nascent electricity grids. While some in the West have apparently decided that delivering energy access is a household event best pursued by developed-world charities, NGOs, and entrepreneurs, what we actually see with electricity is similar to what we see with food — good governance and political institutions are essential for equitable and abundant distribution.
Amory Lovins of the Rocky Mountain Institute weighed in:
On the contrary, everything has changed since 1998, making Marty’s paper even less relevant today. A few quick highlights, doubtless incomplete:
- Half the new generating capacity the world has added since 2008 has been renewable. (In 2013, 68% of China’s new capacity was renewable, the majority of it solar and wind; in Europe, 72%.)
- In each of the past three years, the world invested >$250b investment in non-hydro renewables, adding >80 GW per year. Orders for central thermal plants continue to fade because they have no business case.
- PV power is scaling faster than cellphones worldwide, and in concert with high end-use efficiency (e.g. LED lighting) has extraordinary potential to reach those with no electricity as well as in peri-urban areas.
- China in 2012 increased electricity output more from non-hydro renewables than from all fossil-fueled and nuclear sources, and in 2013, added more PV capacity than the US had added since it invented PVs in 1954. RMI’s Reinventing Fire: China collaboration with ERI, EF/C, and LBNL is turning up cost-effective practical potential to raise China’s 2050 carbon productivity (GDP per unit of fossil fuel) by ≥15x. The potential in India, where a revolution in efficiency and renewables is also emerging, may be even greater. Both these countries make more electricity from windpower than from nuclear power.
- Unsubsidized US market prices in favorable US sites are ~$0.07/kWh for PV power and ~$0.04/kWh for windpower, with both falling rather rapidly, but already they beat new combined-cycle gas, even neglecting gas’s ~$2/GJ price-volatility cost.
- Grid integration of high-renewables mixes, for those paying attention, is now an interesting evolutionary opportunity, not a prohibitive technological or economic challenge. (See http://www.rmi.org/storage_necessity_myth_amory_lovins for a short nontechnical summary of why no breakthrough in bulk electrical storage is needed.) Four EU countries not especially rich in hydropower got about half their 2013 electricity consumption from renewables—Spain 45%, Scotland 46%, Denmark ≥47%, Portugal 58%—with excellent reliability and no additions of bulk storage.
- Micropower—the Economist’s term for renewables, less big hydro, plus cogeneration—now produces one-fourth of the world’s electricity (>2x nuclear output); see RMI’s July 2014 Micropower Database update for details.
- Commonly mentioned renewable issues around land-use turned out to be a canard (http://www.rmi.org/Knowledge-Center/Library/2011-07_RenewableEnergysFootprintMyth).
- Both globally and nationally, renewables excluding big hydro have scaled at least as fast as nuclear power ever did, and without its two-decade “windup” period of building the very demanding capabilities and complex institutions required.
- Properly integrated renewables have emerged as a uniquely profitable and practical pathway to resilient grids that make big cascading blackouts impossible—without adding material cost (see Transform scenario in Reinventing Fire).
- Although U.S. utilities typically pay ~$0.02–0.03/kWh for efficiency, RMI’s empirically grounded Reinventing Fire synthesis showed in detail how to quadruple U.S. electric end-use efficiency at an average cost of ~$0.007/kWh (2009 $). The average IRR for 3-4x higher energy productivity in US buildings is 33%; for doubled energy productivity in US industry, 21%. Previous analyses show higher costs and smaller savings because they left so much out, including integrative design, which often turns diminishing into expanding returns to investments in energy efficiency.
- The world is investing >$300b/y in energy efficiency (the electric fraction is unknown but significant), and some major economies like the US and Germany show pretty steady declines in electricity use even as their economies grow. US weather-adjusted electric intensity, for example, fell 3.4% in 2012 alone. It’s probably reasonable to estimate that efficiency’s annual addition to global electrical services is at least comparable to that of nonhydro renewables.
- It is not essential for developing countries to repeat industrialized countries’ historic trajectories: on the contrary, they can leapfrog in energy supply as many did in cellphones. Some are already doing exactly that. Experience teaches that waiting for the wires to reach the villages, bringing unaffordable thermal power, is impractical. The smart solution is to skip the wires, just as in telecomms, and go efficient/renewable/resilient/distributed.
- In general, developing countries have lower end-use efficiency and have more infrastructure yet unbuilt, and they can more easily build right than fix later, so they have far more dynamic and capacious efficiency opportunities than industrialized countries.
- As Ashok Gadgil and I showed ~23 years ago from World Bank data, investing in negawatts wherever they’re cheaper than megawatts cuts by ~4 orders of magnitude (3 from intensity, 1 from velocity) the capital needed by the power sector — the most capital-intensive sector, gobbling about ¼ of all development capital. Such least-cost investment could turn that sector into a net exporter of capital to fund other development needs. That’s the most powerful macroeconomic lever we know for global development; yet few finance ministers have ever heard of it.
In short, convergent trends in renewable and distributed power, empowered customers, liberalized markets, transparent pricing, and radical end-use efficiency are building an energy future very different from the past. Powerful players are betting on this new horse, coming up fast on the outside. Those who can’t see that horse will continue to lose value, because their preferences have more cost and financial risk than investors wish to fund. These market forces will ultimately prove more important than policy, international agreements, or the inertia of those with old ideas and trapped equity.
I haven’t time to enter a protracted discussion about these facts and ideas, but was surprised by their absence from the thread, and hope their injection may prove helpful.
Chris Green responded to Lovins’s arguments:
It is hard to square the energy facts presented by Amory Lovins with those I have garnered from BP Energy Statistics. While the statistics here (see below) focus on primary energy consumption and those presented by Lovins focus on electricity (a component of final energy) the differences are stark. Here are just a few of the differences:
1. Lovins says renewables (including small hydro and co-generation) account for 47, 58, and 46% of electricity consumption in Denmark, Portugal and Spain, respectively. But non-hydro renewables (NHRs) and all hydro accounted for only 5.7%, 28% and 19%, respectively, of primary energy consumption in 2013 in these three countries. If only NHRs are considered the percentages would be much lower.
2. Globally, NHRs accounted for 2.2% of primary energy consumption in 2013—up from 0.5% in 2000. If all hydro is added in the share in 2013 rises to 8.9%.
3. All of the gain in share of NHRs share since 2000, was offset by the decline in share of nuclear.
4. Lovins claims regarding the contribution of NHRs to electricity consumption in China in 2013 must contend with the fact that the NHRs increased by only 9.4 million million tonnes of oil equivalent (mTOE) between 2012-2013 compared to 100 mTOE for fossil fuels.
As indicated in an earlier intervention, low carbon energy has not made any inroads into the 86.5-87% dominant share of fossil fuels in global energy consumption since 2000. What NHRs have gained has been at the expense of nuclear
I would note that Alex Trembath’s useful intervention to this discussion provides insight into why we can expect global energy consumption will continue to grow, and tangentially why so much of that energy will be supplied by fossil fuels without an major breakthroughs in energy technology.
Trembath built on Green’s comment:
Indeed, Professor Green, the numbers for even 2013 are not nearly as optimistic as Dr. Lovins suggests. Over 2012, 2013 consumption of different energy sources rose by the following amounts, measured in Mtoe (note these are primary energy, not only electricity):
Oil: 46.1
Gas: 34.1
Coal: 103.0
Nuclear: 3.3
Hydro: 22.2
Wind: 24.0
Solar: 6.9
(Source: BP 2014)
There really ought to be a collegial rule against discussing energy consumption trends without mentioning capacity factor, which explain why 68% of capacity added in 2013 was renewable but a much smaller minority of added generation was renewable. Capacity factor of course also explains why renewables excluding large hydro (and excluding CHP) account for almost twice the installed capacity of global nuclear, yet nuclear in 2013 still generated about 20% more electricity than all those sources combined. These graphs are taken from RMI’s Micropower Database; the first is capacity and the second is generation:
Of course we see here also how essential including cogeneration (typically combined heat and power using natural gas) is for drawing such optimistic conclusions about micro generation.
I’m also not sure along what metric Dr. Lovins finds that “ globally and nationally, renewables excluding big hydro have scaled at least as fast as nuclear power ever did.” As this analysis by my colleagues shows, the best cases of nuclear have outpaced the best cases of nuclear when the metric is added generation divided by national population:
Of course some of the fastest energy, not carbon, transitions have been with natural gas. The UK went from 0% natural gas for electricity in 1990 to 40% in 2000. The US deployed over 250 GW of natural gas generation capacity in the last 20 years, and here’s what our own energy transition has looked like:
Certain subsidized and unsubsidized renewables projects can come in at low cost, and we should celebrate these and learn from their successes. But while Dr. Lovins cites 7c/kWh as a best case for solar, the EIA still finds average LCOE of solar at 13c/kWh, higher than advanced nuclear and, of course, higher than natural gas. Wind does appear to have reached an all time low, settling in 4-6c/kWh unsubsidized, as the recent LBNL report showed. However, it must be remembered that these technologies’ value to the grid decreases, and integration cost increases, with higher penetration. This is from the work of Lawrence Berkeley National Labs:
This dynamic is evident in Germany, where wholesale power prices are being depressed by must-dispatch, low-marginal cost renewables, but balancing this intermittency is causing retail power prices to rise, both from increasing FIT commitments, and increasingly with costs like capacity payments for baseload power stations and curtailment payments for excess renewables. Here’s the results of a review of increasing integration costs from variable renewables: