7
Clearly, one of the most compelling public policy arguments for solar PV DG is 8
that it reduces carbon emissions and other pollutants. Expectations of environmental 9
externality benefits may be the biggest motivator for supporting and subsidizing 10
solar PV DG. Proponents of solar PV DG note that solar has zero carbon or other 11
harmful emissions related to the process of producing energy.14 Additionally, solar 12
proponents contend, to the extent that wide deployment of solar PV DG avoids the 13
14 This paper will not enter into the question of whether there are environmental externalities associated with the production of solar panels themselves, as there is no source of energy that, over its full cycle of production, deployment and use, and system impacts, does not have such secondary effects. Solar PV DG panels go through an energy intensive process to be manufactured, much of that occurring in the world’s most carbon intensive energy economy, China, so like every other energy source, solar has a carbon footprint that extends well beyond the carbon-free way in which the panels themselves produce energy.
need to invest in technologies that do have carbon and other undesirable emissions, 1
there is an environmental benefit that comes from avoiding the social costs 2
associated with pollution. In the absence of legal limits on relevant emissions, such 3
costs, solar advocates correctly point out, are not captured in the internalized costs 4
of the competing technologies. Therefore, solar advocates suggest that regulators 5
and policy makers should take these external social costs into consideration in 6
setting prices for various forms of energy.
7
Before delving into this claim any further, it is important to address two issues.
8
First, as referenced above, the idea that regulators or others who set electricity 9
process should use external social costs, as opposed to solely the internalized 10
economics of various forms of energy, as a factor in pricing, is a controversial 11
subject. Many oppose the use of externalities as a factor in pricing because it 12
necessitates social judgments those empowered to set utility rates may not be 13
empowered to make. There is also not always agreement about what should be 14
considered an externality that should be reflected in prices. In the views of such 15
opponents, the only non-internal factors that ought to be incorporated into pricing 16
are those that are internalized by legal mandate. Proponents of incorporating 17
externalities into rates through regulatory action, on the other hand, contend that 18
doing so is the only way to accurately reflect all costs. They also contend that 19
factoring in environmental externalities is a form of insurance against future 20
regulatory requirements. This report does not attempt to take on that large, 21
theoretical/ideological debate about the inclusion of non-mandated externalities in 22
pricing, but rather simply acknowledges that the debate exists and attempts to 23
address the positions advanced by those who favor inclusion of externalities.
1
Without making any judgment about the merits of incorporating externalities into 2
ratemaking, the report will assume, solely for purposes of doing complete analysis, 3
that at least price setters and policymakers might want to consider externalities for 4
purposes of measuring the value of solar PV DG, or, even if they ultimately 5
choose not to do so, they will almost certainly have to confront issues raised by 6
advocates urging them to do so. The report will, therefore, address how 7
externalities ought to be considered and, striving to avoid cherry picking, suggest 8
what externalities might merit consideration in order to make sure that the analysis 9
of the value of solar is both thorough and balanced. Second, it is important to note 10
that the U.S. E.P.A., whose jurisdiction over carbon emissions has been affirmed by 11
the U.S. Supreme Court (Massachusetts v. Environmental Protection Agency, 549 12
US 497 (2007)), has proposed new rules under Section 111(d) of the Environmental 13
Protection Act that would, if promulgated, internalize the costs of carbon into 14
electricity ratemaking. If this were to occur, the issue of whether or not to consider 15
the costs of carbon would no longer be debatable. Thus, in the short term, there is a 16
great deal of uncertainty, which effectively strengthens the hand of those who 17
contend that consideration of carbon emissions would be a form of insurance 18
against future regulation. In the longer run, however, if carbon limitations are 19
imposed, then the cost of carbon will be fully internalized in all energy resources, 20
so special, favorable, arrangements (e.g. pricing) for renewable resources, including 21
solar PV DG, would be of little value in reducing carbon, and perhaps even counter-22
productive to market prices that would already reflect the cost of carbon That, of 23
course, is because in a carbon regulated marketplace, carbon-free resources such as 1
solar PV DG would, ipso facto have higher value on this dimension than carbon 2
emitting resources would have. Should carbon controls be imposed, the market 3
itself should produce the price signals that will inherently reflect the costs of 4
carbon. At that point, special programs such as RPS and RNM could actually serve 5
to impede the most efficient ways of reducing carbon, by diluting price signals that 6
are formed with carbon control fully internalized.
7
Regardless, while acknowledging that, all things being equal, solar PV DG does 8
indeed offer value associated with the absence of carbon emissions and other 9
pollutants, the question the remainder of this section focuses on is whether 10
providing a disproportionate subsidy to solar PV DG (one not available to other 11
renewable or non-emitting sources), is, in fact, helpful in addressing the 12
environmental externality of carbon and other pollutants associated with electricity 13
production, especially (as is the case with SRP), when a renewable preference 14
policy (similar to an RPS) has already been adopted, mandating that 20% of retail 15
energy requirements be met with “sustainable resources” by 2020.15 In fact, as is 16
explained below, disproportionately subsidizing a technology that is not cost 17
effective in reducing carbon emissions, as is true of solar PV DG is, to understate 18
15 According to its website, “SRP has established a goal that by FY20, SRP will meet a target of 20% of its expected retail energy requirements with sustainable resources.” http://www.srpnet.com/environment/renewable.aspx
the point, not a helpful approach to reducing overall carbon emissions from the 1
electricity sector.
2
An environmental analysis of solar PV DG as an emissions reduction technology 3
should include an examination of the least-cost, most efficient ways to get to the 4
desired emission reduction results. As discussed in the first section of this paper, 5
solar PV DG is expensive. Therefore investment in solar PV DG is an expensive 6
way to reduce carbon emissions, based on the levelized cost of energy alone.
7
However, in addition to this, there are other factors that need to be considered in 8
assessing the cost-effectiveness of solar PV DG as an emissions reduction tool, and 9
which raise further questions about the value it offers in this area.
10
There are some characteristics specific to solar PV DG (some of which may apply 11
to utility-scale solar as well) which suggest that a system reliant on this resource 12
may suffer some unintended consequences in terms of increased pollution 13
emissions in other parts of the system. The first such problem may come from the 14
fact (discussed in detail in the next section) that solar PV DG production peaks 15
before the overall system peak, dropping steeply by the time the system peak occurs 16
several hours later.
17
This not only means that solar PV DG has less value as energy, but it is also 18
powerful evidence that it also has less environmental value. This is because, as a 19
general rule, the least efficient and “higher emitting” plants are dispatched at times 20
of peak demand. Thus, when solar PV DG is producing energy, it is not displacing 21
the most carbon emitting plants. Instead, it is displacing more efficient, less 1
polluting generating units.
2
Second, solar PV DG is an intermittent and unpredictable resource. As such, its 3
availability is uncertain and in order to supplement it, fossil plants are often called 4
upon to operate on a less efficient, more carbon-emitting "ramping" basis than if 5
they were running as pure baseload.
6
The intermittent nature of solar PV DG16 has still another effect that may serve to 7
dampen environmental expectations. Because the capacity required to supplement 8
the renewable is ramping rather than baseload (and thus able to sell electricity for 9
many fewer hours), the signals to investors to build new, more efficient generators 10
is diluted, and is therefore less attractive from a financial point of view. (This is a 11
particularly interesting issue in the context of the California duck chart.) If the new 12
investments are not made, then older and likely “higher emitting” plants will have 13
to have their lives extended and/or be operated on a ramping basis for which they 14
were not designed. The result will be less efficiency and, therefore, likely increased 15
emissions.
16
16 Intermittency is also an issue for wind and large scale solar, but those two resources have the advantage of being less expensive, especially so when NEM is the basis for pricing solar PV DC, so their cost effectiveness in reducing carbon is much better than solar PV DG.
Thus, to be truly meaningful and intellectually honest, any analysis of 1
environmental impact must take into account the change in dispatch and operations 2
Among analyses that take these kinds of factors into consideration, there is 3
agreement that it is the least efficient of all renewable energy resources (including 4
demand side/energy efficiency measures) in common use in this country. An 5
interesting dialogue occurred recently between Charles Frank, an economist at 6
Brookings, and Amory Lovins of the Rocky Mountain Institute, based on an effort 7
by Frank to develop an analysis of the cost-effectiveness of solar PV as a carbon 8
reduction tool, taking into account not only the levelized cost of energy, but some of 9
the considerations about peak production and effects on the functioning of the 10
overall energy system discussed above.17 Their dialogue, while contentious on many 11
points, includes, on both sides, numbers that would suggest agreement on the fact 12
that solar PV is the least cost effective of all carbon-benign resources in reducing 13
emissions. Frank analyzed five generation resources by their cost effectiveness in 14
reducing carbon and concluded that nuclear and natural gas, followed by hydro, 15
17 See Charles R. Frank, Jr., “The Net Benefits of Low and No-Carbon Electricity Technologies.” Global Economy and Development at Brookings Working Paper 73, May 2014; and Amory Lovins, “An initial critique of Dr. Charles R. Frank, Jr.’s working paper ‘The Net Benefits of Low and No-Carbon Electricity Technologies,’
summarized in The Economist as ‘Free exchange: Sun, wind and drain’.” Rocky Mountain Institute, 2014. (http://www.rmi.org/Knowledge-Center/Library/2014-21_Frank-Rebuttal)
wind, and solar were, in that order, the most cost-effective types of generators for 1
reducing carbon.
2
Lovins took issue with Frank for using outdated data and for not looking at energy 3
efficiency. After a colleague re-ran Frank’s calculations using Lovins’ suggested 4
data, Lovins reported the following:
5
Instead of gas combined-cycle and nuclear plants’ offering the greatest net benefit 6
from displacing coal plants, followed by hydro, wind, and last of all solar, the ranks 7
reversed. The new, correct, story: first hydro (on his purely economic assumptions) 8
and gas (only if we omit its price volatility), then wind, solar, and last of all 9
nuclear—still omitting efficiency, which beats them all.18 10
Here, he shows nuclear ranked last in cost effectiveness, and it should be noted that 11
he expresses some reservations about the ranking of natural gas, because of its price 12
volatility, which is not considered in the rankings. What is significant, however, is 13
that among renewable resources, as noted above, solar continues to rank as the 14
least cost-effective renewable resource for reducing carbon, even in Lovins’
15
analysis.
19
Lovins himself does suggest that this may not be his final word on the 16
18 Lovins, Amory B. “Sowing Confusion about Renewable Energy.” Forbes August 5, 2014.
19 Charles Frank blog post, “ Alternative Energies Debate—The Net Benefits of Low and No-Carbon Electricity Technologies: Better Numbers, Same Conclusions”
September 4, 2014. http://www.brookings.edu/blogs/planetpolicy/posts/2014/09/04-low-carbon-tech-lovins-response-frank
As Frank puts it, even after addressing Lovins' criticisms, "Wind continues to rank number four and solar ranks number five by a large margin."
subject—he notes that the analysis would have more value if “other hidden costs, 1
risks, and benefits” were counted. But it remains significant that according to the 2
analyses of both men -- who hold quite divergent views on how best to reduce 3
carbon emissions -- not only is solar PV expensive, it is the least cost-effective 4
renewable resource for reducing carbon emissions. Solar PV DG itself was not 5
considered in these calculations—however, based on the Lazard pricing analysis 6
shown at the beginning of this paper, it is safe to assume that the economics of 7
distributed solar as a carbon reduction mechanism are even worse than the numbers 8
for utility-scale solar PV.
9
The lack of direct correlation between increasing the amount of solar PV DG in an 10
electric system and reducing carbon emissions has been dramatically demonstrated 11
by developments in Germany. In Germany, where there has been a very dramatic 12
increase in reliance on intermittent energy resources, prices have risen since 2005.
13
That was not surprising, but what was not expected was the spike in carbon 14
emissions that resulted (See Eddy, Melissa, “German Energy Push Runs Into 15
Problems.” The New York Times, March 19, 2014). While there are very significant 16
differences between the SRP territory and Germany, (perhaps most notably that 17
Germany has been phasing out its nuclear fleet), the experience in that country is 18
very telling. It clearly demonstrates that an increased dependence on renewable 19
energy resources, and particularly intermittent resources, does not, as many solar 20
proponents claim, ipso facto mean fewer carbon emissions, and may, in fact, cause 21
the opposite to occur. It also demonstrates that prices can escalate dramatically 1
when subsidy mechanisms (in the German case, feed-in tariffs) are far in excess of 2
market prices. The Germans, incidentally, have recognized their miscalculations 3
and are in the process of recalibrating their strategy.
4
Problematically for the environmental “value of solar” argument, attempts to reflect 5
the “value of solar” in preferential pricing of DG PV, as opposed to non-DG forms 6
of renewables, may lead to distortions that favor DG PV over larger-scale, usually 7
more efficient and less costly forms of non-emitting generation that will achieve 8
more environmental benefits at lower cost. Results such as that cannot be justified 9
on the basis of externalities, which are no different between DG PV and larger-scale 10
renewables. Indeed, it could well be argued that overpayment for DG PV would 11
have the effect of squeezing more efficient forms of renewable energy out of RPS 12
markets by using preferential pricing to grab a disproportionate share of the RPS 13
market for a less efficient resource. In the long run, of course, the inherent 14
favoritism in pricing PV DG over other renewable energy sources does not bode 15
well for the future of renewables. Discrimination in favor of inefficient resources on 16
a long-term basis is almost never sustainable. The inevitable backlash in the 17
marketplace has the potential to sweep away public support for renewable energy in 18
general, an outcome no one concerned about the environment would want.
19