Further falling solar PV power cost also in the long run!   How can countries make use of this encouraging prospect? 

Agora Energiewende recently published  a study commissioned to the FhG-ISE  on ‘Current and Future Cost of Photovoltaics’, which corroborates the rationale for further falling cost of power generation from solar PV.  On the basis of the study results Agora Energiewende calls on countries with good irradiation to improve the financial and regulatory environment and to reassess the role of solar PV in scenarios of future energy systems.  These messages from a German organization addressing other countries’ policies merit a closer look and appreciation:  Is the basic study methodologically sound and technically solid? And are these policy calls well founded? Is the levelized unit cost concept approptriate for economic evaluation of power generation technologies? How can cost assessments serve in the decision support systems for long term strategy and shorter term expansion of electricity systems, in particular for developing countries?

The brief elevator pitch “Key insights at a glance” presented upfront in the publication from February 2015 says: “

  1. Solar photovoltaics is already today a low-cost renewable energy technology. Cost of power from large scale photovoltaic installations in Germany fell from over 40 ct/kWh in 2005 to 9ct/kWh in 2014. Even lower prices have been reported in sunnier regions of the world, since a major share of cost components is traded on global markets.
  2. Solar power will soon be the cheapest form of electricity in many regions of the world. Even in conservative scenarios and assuming no major technological breakthroughs, an end to cost reduction is not in sight. Depending on annual sunshine, power cost of 4-6 ct/kWh are expected by 2025, reaching 2-4 ct/kWh by 2050 (conservative estimate).
  3. Financial and regulatory environments will be key to reducing cost in the future. Cost of hardware sourced from global markets will decrease irrespective of local conditions. However, inadequate regulatory regimes may increase cost of power by up to 50 percent through higher cost of finance. This may even overcompensate the effect of better local solar resources.
  4.  Most scenarios fundamentally underestimate the role of solar power in future energy systems. Based on outdated cost estimates, most scenarios modeling future domestic, regional or global power systems foresee only a small contribution of solar power. The results of our analysis indicate that a fundamental review of cost-optimal power system pathways is necessary.”

These points are somewhat elaborated in the longer version of ‘key insights’  presented as chapter 1 of the publication. Scanning the next 7 chapters, which form the study proper,  the reader notices that the ‘key insights’ are not all directly drawn from the actual study.   

In the analytical body of the study, introduced in chapter 2 and conducted by the Fraunhofer-Institut für Solare Energiesysteme (FhG-ISE),  one finds:

a)  a very elaborate assessments of future  cost development of the various components of PV power generation including modules, inverters and others, based on broad expertise and technical analysis (chapter 3, 4 and 5),

b) the calculations of Levelized Unit Cost of Electricity (LCOE) from PV for installations made in the future up to 2050, based on the assessments  of the components’ cost  and further assumptions (chapter 6), which leads to the extremely low LCOE mentioned in the key insights,

c) a Sensitivity analysis for varying cost of capital (also in chapter 6), and a discussion of results (chapter 7) , and finally

d) uncommented calculations of LCOE from PV for a number of OECD and  developing countries and their respective range of solar irradiation as well as capital cost assumptions (in chapter 8).

Thus, the study has a more limited scope vis-a-vis the ‘key insights’. It does not refer to the past cost reductions (key insight 1). It does not discuss ‘financial and regulatory environments’ (key insight 3) and it provides neither overview nor reference to an overview of future power system scenarios (key insight 4).

The study does however provide a quite solid analysis and assessment of future components cost developments and translates that into LCOE from PV at different points in future time. Thus, the good news of key insight 2 seem well  founded. 


Study results and political message

It seems, however, that the FhG-ISE has restrained the content of statements to what can be concluded from the analysis, which is what scientists do. In contrast, Agora Energiewende has woven a narrative around the core study. First it embeds it within  the recent past of price developments,  current price levels and eeed-in tariffs to formulate key insight 1.  On the basis of the study, it implies, that it is the ‘financial and regulatory environment’ which  impacts heavily on the cost of capital the parameter (WACC) used in the study to demonstrate the cost sensitivity,  in order to call for improvement in key insight 3. Finally,  in key insight 4,  Agora Energiewende asserts that  ‘Most scenarios fundamentally underestimate the role of solar power in future energy system’ and suggests that work should be done to revise these on the basis of the study results.  In the study itself there is neither an overview  nor reference to an overview of power system scenarios and role of solar PV in scenarios. 

Thus, in the interest of politically useful messages, the Agora Energiewende carries the conclusions far further than the study itself provides.  The statements made may be true, and most of us agree from cognizance and experience, that financial and regulatory environments impact cost of capital and that most power sector scenarios underestimate the role of solar power, but theses statements are not directly supported by the study.  They were, apparently, common sense of the expert groups which contributed to and discussed the study assumptions and results, as mentioned in chapter 7. However, these differences between the content of  study itself and the ‘insights’ drawn from it call our attention to the auto-denominated ‘think tank’ and its role.

The Agora Energiewende (find out more) is designed, to my understanding, as a market place of ideas in the sense of the agora in ancient Greece.  In order to feed and enliven this virtual market, it provides studies, publications and information to the public, and organizes discussions. It assures its independence from vested interest through financing of their operations and of the committed studies from foundations. Its supervising council has a broad scope of representatives from society including the power industry and their clients.  The common denominator which makes this possible  is the interest to make the German Energiewende a success, which is shared by all political parties and a major share of the German population. Actually, the current State Secretary in charge of Energy  in the Federal Government was the founding  Director of the Agora Energiewende.  Since in office the newly constituted Federal Ministry for Economy and Energy has established similar mechanisms  of developing and discussing policies based on  studies, in particular the Plattform Strommarkt and the Green Paper process.

It is no pettifoggery  insisting  that there are marked differences between the conclusions published by Agora Energiewende and the body of the study.  It seems (always) important to note such difference, in order to appreciate carefully the foundation of political messages which are supposedly underpinned by research.  This distinction is also important for the interpretation of falling LCOE of PV and what can be drawn from this result  by developing countries.


A brief discussion of LCOE, past and present 

LCOE have been a mainstay of economic comparisons of different power generation technologies since decades.  The somewhat simplified rationale used to be, that electricity demand and its profile over time is a given, so the economic evaluation could concentrate on the cost. From the long term projection of demand and available capacity one could identify blocks of  capacity and energy requirements, specifically a  'base',  'intermediate' or 'peak load' demand to be covered.  Based on the assumption, that the generation technology for the new capacity would not have an impact on revenues and other system cost, the  economic comparison and calculation could  be limited to a cost comparison (instead of a cost-benefit calculation).  The frequent cost comparisons since the late 1970s of nuclear and coal power technologies for base load in the US conducted by the Energy Information Agency (EIA) and in Germany by my colleagues from the Institut for Energy Economics at the Cologne University are the classical examples and were used in the political battle for and against nuclear, and to my recollection were the occasion for which the LCOE were developed.  The question to be answered was basically: ‘What is the unit electricity price for the power produced  over the lifetime of the plant which would cover all cost caused by construction, operation and decommissioning'. The methodology would also allow to internalize external cost and even include all social cost, in other words (calculative), as several  works[1]  since the 1980s and recent overviews demonstrate.

At the time, this comparison based purely on cost seemed more or less admissible, in cases of technologies with similar characteristics, e.g. comparing inflexible plants of high availability or at same locations, which could be situated at similar integration cost within the transmission grid  etc.  Thus, the cost-based economic comparison of nuclear, lignite and hard coal plants e.g.  in Germany was accepted, although  the flexibility of hard coal and nuclear plants are not the same, and the lignite plants (in Germany) were bound  to mine-mouth, whereas the nuclear and hard coal plants needed  locations at rivers.  In addition, the high concentration of lignite fuelled capacities in a region required transmission lines there and was connected to (pump) storage hydropower plants in the alps, to accommodate such  large inflexible base load capacities.  Thus, the technologies   involved  specific additional transmission and interchange cost, which made it less appropriate to work with cost comparisons, even when including  such cost in the calculation.  The assumption of a given demand profile was not quite correct, since the high availability and low flexibility of nuclear and lignite power plants  led  not only to build more storage hydropower; it also led  to intensive marketing for new off peak electricity demand, namely (night-storage) heating  in Germany.  This new demand was  less valuable and the energy could only  be sold at lower prices and required  controls.

The reader may already see some parallels to the current  economic evaluation  of power generation.   With the advent of renewable energy power generation technologies based on fluctuating resource availability like solar and wind, the question arises anew, whether LCOE methodology can be used for the economic evaluation in view of the variability of generation. 

Similarly, the ranges of LCOE for a given technology may be used in an assessment of economically and technically proved resources of vRE and potential supply from a specific power generation technology in a given country. In this respect, the results of the present study more than confirm the large potentials classified by cost range for solar PV in a 2008 REN21 study with DLR and ECOFYS -  to my own relief, since I was responsable for that study. 

Obviously, there is little concern inn using LCOE, when we use it to calculate electricity generation cost of a technology  or source without comparing it to another technology, e.g. comparing the cost at different points in time or in different locations for the same technology. This is exactly what the present study  does.  So, the methodology is entirely correct for the study itself.  

The authors of the study also maintain correctly in section 6.2 that the LCOE methodology is not a sufficient basis for an investment decision  in a specific plant.  It may serve, however, as first orientation for potential investors who can compare on the basis of LCOE of the plant in question the required  price of energy  with the feed-in price  for the energy delivered  in case of  Feed-in-Tariff rule,  or with the  price of energy not drawn from the grid, i.e. the power tariffs  in case of Net-Metering  rule.  In any case, LCOE has limited value for individual investment decision making.  

In the present study, there is no LCOE based cost comparison of solar PV power with other generation technologies.  So we can keep the discussion short, whether LCOE may be used in comparisons of  different generation technologies with distinct characteristics of availability, such as variable Renewable Energy  (vRE) power from wind and solar on the one and dispatchable fossil, bioenergy  geothermal or inflexible nuclear power on the other side.  It is obvious, that  the comparison of  plain LCOE of vRE and non-vRE, although frequently carried out, e.g.  by OECD and IEA as well as annually by the EIA  is more problematic  than the comparison of  LCOE of coal and nuclear power, and needs to interpreted with care, and some aspects outside of the plant itself should be included.  As veteran energy economist Joskow writes here:  “Integrating differences in production profiles, the associated variations in the market value of the electricity at  the times it is supplied, and the expected life-cycle costs   associated with different generation  technologies is  necessary to provide meaningful economic comparisons between  them”.  One might add, that also grid cost differences associated to the technologies need to be considered.  In view of the upcoming new low carbon power system paradigm with distributed generation, fluctuating RE based generation, storage and smart grids as well as demand management and new users like electro-mobility  it seems  appropriate to  consider the  whole system  in economic evaluation of future roles of generation technologies.


The necessary  review of decision support systems

Key insight 4 of the present Agora Energiewende publications, concludes: “The results of our analysis indicate that  a  fundamental  review of cost-optimal power system pathways is necessary. 

Actually, this and also the key insight 3, are not very strong request. Agora Energiewende does not suggest that,  in view of the results, there should be an all-out investment from n ow on in solar PV in the worlds sun-belt countries. Such request would be met by serious incomprehension and even suspicion, in particular by experts and politicians from develooing countries with ever growing power demand and intentions to enlarge the industrial value chain and to establish processing industries. 

There is an implicit understanding, that low LCOE from solar PV are only part of the rationale. The other important parts consist in the tasks of accomodating the fluctuating and sometimes very low power supply from PV in systems characterized by a rather recurring daily and nightly demand pattern. And as said in a paragraph  above, numerous complementary and flexibility options can and must contribute to that, the cost implications of which are not figured out in the present study.

Thus, focusing on scenarios like the Agora Energiewende does in key insight 4, is a reasonable step. When this request would be fulfilled, that would produce scenarios in which the role of solar power in future energy systens would be more significant. So what? one  may ask, since scenarios are not reality.  What is important is that decisions are taken on the basis of scenarios and other decision support to drive future reality towards the desired scenario, taking into account the new options which may appear on the way. Let us look a little closer to scenarios and more generally to decison support tools in which the updated PV cost information should be integrated.

 Long term power sector scenarios are designed according to different narratives of future development in the respective country.  In most scenario studies, one straightforward business as usual scenario is a reference case, and is contrasted to distinct scenarios of technological, economic and social evolution from now. Or, scenarios might be designed backwards, responding to the question:  Given a desired long term future, what potential developments would connect it to the current situation?  In either case, One high solar PV scenario would be desirable to be included.

In view of the increasingly complex tasks of sufficient and secure electricity supply including high shares of vRE,  it is extremely important to take not only annual demand and  the annual load duration curve  as a demand  data the future year but also some typical and critical daily load and supply curves,  in order to study the flexibility options   from  import- and export,  storage and grid elasticity,  distributed generation and demand shifting, efficiency as well as serving additional uses like electro mobility.   A good example for  how to include this is in the  Long-term scenarios and strategies study  by DLR/FhG-IWES and IFNE for the German Environment Ministry.  

In the next step, the evaluation of the scenarios, the economic cost – benefit implications would be assessed, in addition to security of supply,  GHG emissions  and other criteria derived from political objectives.  Here the new and downsized cost estimates for the components  solar PV  component would be included, and system cost of the different scenarios could calculated, typically total investment requirements,  operating cost,  total cost   not for a single technology but for the system as a whole. The comparison of scenarios would  show that the solar scenario would be more advantages than thought  before, taking into account all system complements.

Long term scenarios are useful as guidance for the choices of strategies and  pathways. This is particularly important in power systems which consist of long lasting infrastructure, in order to avoid a wrong turn and technology lock-in.  This may lead to considerations of preferring technologies which currently are a bit more expensive but have the prospect of  becoming clearly cost effective and help restructuring the power system away from centralization to distributed generation and  to cost effective grids. 

The scenarios, though,  may give orientation but do not tell what to do next and how to decide in a specific current choice situation.  Actually, for short and medium term power system decisions current and near future technologies and cost are obviously relevant.  Developing countries, though, do need generation capacity expansion  fast apart from improvements in supply quality and reliability.  Solar PV capacity  can be expanded rather quickly as well as wind (which is more location bound than solar),  and they cannot wait for lower solar PV cost to materialize.

It has,  not quite incidentally,  been demonstrated for several developing countries by a BMZ-sponsored activity of GIZ, that already now Operational Benefits of Growing PV and Wind Shares in National Power Grids are such that more solar and wind power can be incorporated cost-efficiently. The  falling  solar PV cost makes the scale-up within the given system increasingly more economic,  as a comparison of LCOE from solar  PV and the respective unit value of  operational benefits reveals.

Thus, there is effectively a  need for review, as Agora Energiewende suggests, although the review should involve the whole decision support tools for short, medium and long term power sector development. The long term scenarios modified in particular with respect to the solar PV cost would show the preferred pathway, whereas the short and medium term power system models would indicate the options for fast integration of vRE  in the short and medium run and would need to be coherent with the long term pathway.

[1] The pioneering work of O. Hohmeyer ‘ Social Costs of Energy Consumption - External Effects of Electricity Generation in the Federal Republic of Germany’  published in 1988 by Springer merits mentioning; http://link.springer.com/book/10.1007%2F978-3-642-83499-8


Write a comment

Comments: 0



Dr. Paul H. Suding                Mail: paul@elsud.net

Greijze Graaf 9                    

6245 KG Eijsden/Mesch 

The Netherlands