Draft – Sep 13, 2005 Valuing Renewable and Conventional Generating Assets in an
Environment of Uncertainty and Technological Change
Environmental Audit Committee, House of Commons
New Enquiry: Keeping the Lights on: Nuclear,
Renewables and Climate Change
Shimon Awerbuch, Ph.D.
SPRU Energy Group ? University of Sussex
Electricity capacity expansion questions currently focus on two principal issues:
i) The kilowatt-hour cost of renewables such as wind, relative to the cost 1of gas turbines and, potentially, nuclear power; and
ii) How our century-old electricity grid might be contorted into dealing
with the so-called intermittency of wind and other renewable
In this submission I argue two principal points. First, that while electricity
planning has long relied on the stand-alone generating costs of various technologies, this measure is no longer relevant. In its place, I suggest that
electricity policy be based on modern portfolio theory concepts, which reflect the cost as well as the risk contribution a given generating technology makes to
the generating mix. Financial investors routinely use portfolio optimization
techniques to value stocks and other additions to their holdings. These
techniques consistently show that when added to a conventional generating mix,
wind and other fixed-cost renewables serve to lower overall generating costs.
This outcome, which is predicted by finance theory, holds even if it is assumed
that the stand-alone costs of wind exceed those of gas.
Second, I argue that the current debate about the system integration costs of
wind and other variable output renewables is largely misplaced. This debate
conceives of wind as a direct substitute for dispatchable fossil technologies,
which it is not. As a consequence, the debate needlessly dwells on such issues
as the cost of additional backup generating capacity. In my opinion, efficiently
integrating wind and other new, passive, variable-output renewables will
ultimately require changes our current electricity production-delivery paradigms
and protocols. This is a tall order. At the very least, integration will require
new parallel in formation networks for the electricity grid and most likely new
1 This is not to say that other issues, such nuclear waste are not also prominent.
? Shimon Awerbuch, 2005 1
– Sep 13, 2005 Draft ways of charging for grid services. These must allow wind-based electricity
products such as space and water heating, which naturally match this
‗intermittent‘ technology with so-called ‗dispatchable‘ load applications. The
century-old concept of the grid as a system for transporting commodity electrons
becomes obsolete in an environment characterized by many distributed
generating sources and a diversity of load applications.
I. The cost of renewable and conventional technologies
Intelligent electricity policy and planning decisions cannot be made on the basis
of traditional engineering kilowatt-hour (kWh) cost models. Such models,
developed around the time of the Model-T Ford, have been widely discarded in
most industries in favour of modern asset valuation concepts.
My own stand-alone kWh cost estimate, which have remained quite constant
over the last decade, use a standard finance-oriented (CAPM) approach that
reflects the cost of risk. Through periods of both low and high fossil prices and 2money-market rates, these estimates have consistently suggested that gas generation costs more than wind and many other renewables (e.g. Awerbuch
June 2004, June 2003, May 2003, February 2003, November 2000, April 1995,
1993). This is in stark contrast to estimates prepared by the DTI (2003), the
Royal Academy of Engineering, the IEA (2005) and other national and
international agencies, which generally find that gas costs less than wind. These
agencies however, use calculative procedures that produce cost results with no
economic interpretation; they should not be given any weight in policy-making.
Fossil fuel prices have risen significantly over the last two years and some
predict oil-price spikes in excess of $100/bbl (Reuters, 2005). Finance theory
tells us that future cost streams can be meaningfully expressed only in terms of
their market risk. When fuel price volatility is considered, gas-based generation
is noticeably more costly than standard engineering-based estimates indicate.
For example, conventional estimates such as those produced by DTI and IEA
suggest that gas-based electricity costs in the range of ?0.03–?0.05/kWh. The true, risk-adjusted cost is quite likely in excess of ?0.06–?0.07/kWh, making 3many renewables generally competitive.
II. Externality Benefits of Wind: Enhancing Cost and Energy Security
Risk-adjusted CAPM procedures more reliably estimate the cost of gas and wind.
But even CAPM results are only as good as the underlying gas price forecasts,
which could dramatically change even before this inquiry is concluded.
Meaningful kWh cost estimates require unbiased gas price forecasts. But
2 Inflation expectations and money-market rates underlie CAPM-based discount rates.
3 The cost advantage of wind survives when system integration charges are added, e.g.
per Dale, et. al. (2004) or the DENA Grid Study (2005).
? Shimon Awerbuch, 2005 2
– Sep 13, 2005 Draft
history provides little comfort that today‘s fossil price forecasts will be any
more reliable than those of the past. Nonetheless, suppose we assume for the
moment that the conventional analyses— which predict that gas generation costs less— are correct. Does this imply that we should abandon wind and other
options and invest only in gas? This is the traditional ―least-cost‖ approach to
energy investment decision-making. It makes little sense in today‘s highly
uncertain energy environment.
Energy planners need follow financial investors, who are used to dealing with
risk. No one can predict the performance of a corporate share of stock over 20
years just as no one can predict the price of gas. Financial investors deal with
market risk by holding efficient, diversified portfolios. These offer the best
hedge against an uncertain future. Prudent investors do not try to chase today‘s
best performing securities; these may be the laggards of tomorrow. Rather, they
balance their portfolios with a mixture of potentially high yielding securities
along with low-yielding government obligations and similar ―safe‖ investments.
Policy makers must take note of this important idea. It matters little that gas
might appear to be the lowest cost alternative (on the basis of conventional
costing models). Even if correct today, that picture could change dramatically,
suggesting that electricity planning and energy policy making in general must
abandon its fixation with identifying alternatives with the lowest stand-alone
cost and focus instead on developing optimal generating portfolios and
When added to a risky, fossil-dominated generating mix, wind and other fixed-
cost renewables reduce generating cost and risk, as long as the mix can be
altered over time [Awerbuch 2005, February 2000, 1995, Awerbuch and Berger
2003]. This so-called portfolio effect, (e.g. see Brealey and Myers, 2003) holds
even if wind costs more on a stand-alone basis. Wind‘s generating costs are
uncorrelated to fossil costs which means that it diversifies the mix and reduces
expected overall cost and risk the same way diversification improves the
expected performance of financial portfolios.
For example, DTI‘s Year-2010 target generating mix (DTI, 2004, 2003) has an
overall cost of 2.96 p/kWh (Table 1). It consists of 71% fossil generation and
11% wind. By contrast, applying the same generating costs, it is possible to 4 which cost no more, but have identify a number of optimized UK portfolios
wind shares of 31% to as much as 54%— three to five times as much wind as the DTI target mix. The ‗Equal Cost‘ portfolio, (Table 1) has the same cost but
lower risk than the DTI target mix, yet contains 31% offshore wind, in spite of 5the fact that this technology is assumed to cost 75% more than gas.
4 An infinite number of such portfolios exist, all with different cost-risk and different
technology shares. 5 This study focuses on wind. Nuclear output is constrained so it does not exceed 2004
? Shimon Awerbuch, 2005 3
– Sep 13, 2005 Draft
DTI Targets Versus Optimized Generating Portfolios
(UK – 2010)
DTI 2000 / 2010 Technology Generating Costs (p/kWh):
Coal: 4.0 / 3.6 - Gas: 2.0 / 1.9 - Wind: 2.7 / 2.0 - Offshore: – / 3.6
Typical Optimized Portfolios DTI 2010
Target Portfolio ‘Equal Cost’ ‘Equal Risk’
2.96 p/kWh 2.96 p/kWh 2.49 p/kWh Portfolio Cost
Portfolio Risk .08 .04 .08
Fossil Share 71% 32% 52%
Nuclear Share 16% 12% 14%
On-shore: 11% On-shore: 25% On-shore: 31% Wind Share
Offshore: 0% Offshore: 31% Offshore: 0% Source: Awerbuch Airtricity (2005)
These results are not meant to suggest that 50% wind shares are feasible given
today‘s network architecture, or even that such a target cold be attained in five years. The results are presented to illustrate that stand-alone costs, even if
adjusted for risk, are not necessarily a meaningful metric for evaluating energy
options. Because various technology costs move in unison, (e.g. are correlated),
intelligent energy strategy, by necessity, requires that cost interrelationships be
considered. Electricity capacity planning must reflect the cost and risk of the
The UK results shown above are representative of similar, and in some ways
even stronger results I have obtained for the US, the EU, as well as Mexico,
Morocco, and other nations (Awerbuch, 2005, Awerbuch, Jansen and Beurskens,
2003). The portfolio approach illustrates the idea that increasing the
deployment of wind, even if it is assumed to cost more, does not necessarily raise overall generating cost, as long as the generating mix can be re-optimized
over time. Wind production costs are relatively fixed. This creates important
cost-risk benefits for generating portfolios.
Energy Security – The Oil-GDP Effect
Oil price increases and volatility dampen macroeconomic growth by raising
inflation and unemployment and by depressing the value of financial and other
assets. This so-called Oil-GDP effect has been reported in the academic
literature for a quarter of a century, although it received little attention from the
media and energy policy makers prior to the recent oil price spikes. The Oil-
? Shimon Awerbuch, 2005 4
Draft – Sep 13, 2005
GDP effect is sizeable. In a recent paper Raphael Sauter and I (2005) estimate
that a 10 percentage-point increase in the global share of wind (or other
renewables) can help avoid GDP losses of $95–$176 billion (Table 2).
Avoided GDP Losses for 10-Percentage-point Increase
in the Wind Generation Share (USD $Billions)
US EU-15 OECD World
2003 GDP $10,882 $10,970 $18,659 $36,356
Avoided GDP Losses
High Estimate $53 $53 $90 $176
Low Estimate $29 $29 $49 $95
Source: Awerbuch and Sauter, 2005 Table 2: Wind Deployment Offsets Sizeable Macroeconomic Oil-GDP Losses
These avoided losses offset 20% of the renewables investment needed to meet
2020 EU RES-E targets and 40% the OECD requirements. Our analysis
suggests that each additional kW of wind helps society avoid $250 in GDP
losses. Stated differently, avoided GDP losses offset 20%-25% of today‘s investments in wind and other renewables.
Energy security is enhanced when nations hold optimal generating mixes that
minimize exposure to fossil volatility. As the last two sections have described,
wind and other renewables provide a joint set of benefits: they enhance energy
diversity/security while they reduce overall generating costs (Awerbuch, Stirling,
Jansen and Beurskens 2006).
stIII. Modernizing Power Networks to Accommodate 21 Century
6Widespread debate prevails about how to manage wind‘s variable-output and how to make it fit into today‘s electricity production-delivery system, designed
over a century ago for dispatchable, fossil-fired central station generation. Had
a different generating technology emerged in the 1890‘s, it would have no doubt
given rise to a different set of network system architecture and protocols. But
we are stuck with our system– at least for time being. System engineers have
been weaned on dispatchable technologies with central control. It is difficult for
them to imagine anything else: they see the challenge as making wind fit into
6 The concept of wind intermittency is misleading. Wind blows a high percentage of
the year, at least at better sites, although its force varies so that output is variable.
There are very few days when wind ceases entirely implying that variable-output is a better concept.
? Shimon Awerbuch, 2005 5
– Sep 13, 2005 Draft
the system. I see the challenge as rearranging the electricity production-delivery stparadigm to accommodate a variety of 21 Century needs, including the
integration of wind and other variable-output sources.
Many new process technologies have faced significant impediments to their
integration and were fully exploited only after underlying systems and
infrastructures were extensively modified. We tend to conceptualize new
technologies in terms of the capabilities and functionalities of the previous
vintages that we better understand. This was true for word processing, which
was initially conceived merely as a replacement for the typewriter, and is true
for wind. How do engineers want to integrate wind? By making it act like a gas
turbine (as much as possible) so it can be centrally dispatched by the control
room operator, just the way it has been done for a nearly a century. Fully
integrating wind will likely require new approaches, including different system
architecture and protocols and powerful parallel information networks to
manage electricity grids in a decentralized, market-responsive manner.
We need to alter the electricity production-delivery system to better staccommodate 21 century needs and capabilities. This involves adopting mass-
customization concepts from manufacturing and moving decision making to loads, which have better information about their hour-to-hour requirements than
a central dispatcher. At any moment, the system‘s total load consists of
thousands of transactions, each with a different value. Electricity to power
water pumping or heating likely does not have the same value as electricity
required for microchip processing (Awerbuch, March 2004, July-August 2004).
Adapting to these realities will yield a more efficient, more market-oriented
production-delivery paradigm under which the network operator becomes the
electricity market enabler. The traditional transportation function of the network
becomes obsolete in an environment characterized by a large number of
Today‘s network is based on outmoded mass-production concepts. Electricity
mass customization will allow users to take power in the forms that best match their various applications. Implementing these ideas requires new strategies for
regulating network system operators, who hold a key position in an electricity
system that has been partially deregulated in the belief that markets, not
regulation, produce the greatest efficiency (Awerbuch, Hyman, Vesey, 1999,
Chap 3). Yet the system operator continues as a monopoly entity with no
incentives to create new market-driven products or to diversify the mix to
broaden consumer access to competitively priced supply markets that include
traditional generation along with wind and other renewables.
Policy-makers correctly focused on deregulating generation first. Much of the
potential benefit of those policies however is lost because the essential market
facilitator, the transmission operator, is naively conceived as a caretaker of the
wires with no incentive to enhance overall system performance. Efficient
integration and exploitation of wind may have to wait until policy-makers focus
? Shimon Awerbuch, 2005 6
Draft – Sep 13, 2005
on the governance, organization, regulation and pricing structures of electricity
CAPM-based risk-adjusted procedures suggest that at currently projected gas
prices, wind and other fixed cost renewables are likely to provide electricity at
lower cost. Moreover, when added to fossil-dominated generating mixes, fixed-
cost renewable technologies reduce cost at any level of risk by virtue of the
portfolio-effect. This holds even if they are assumed to cost more on a stand-alone basis.
Wind and similar fixed-cost technologies enhance energy security and their
deployment will help the UK avoid costly macroeconomic (GDP) consequences
induced by oil price volatility. Every kW investment in wind offsets $250 USD
in oil-induced GDP losses. The benefits of wind and other renewables are
strong, verifiable and highly certain. Our challenge is to re-engineer the
electricity production-delivery paradigm so it efficiently integrates variable-stoutput renewables and meets other 21 century requirements.
Awerbuch, S. (2005), ―Portfolio-Based Electricity Generation Planning: Policy
Implications for Renewables and Energy Security,‖ Mitigation and Adaptation Strategies for Climate Change, in-press
_______ (July-August, 2004), Restructuring Electricity Networks: decentralization,
mass-customization and intermittency, Cogeneration and On-Site Power Production _______ (June, 2004), ―Towards A Finance-Oriented Valuation of Conventional and
Renewable Energy Sources in Ireland,‖ Dublin: Sustainable Energy Ireland, June www.sei.ie/uploads/documents/upload/publications/shimon_awerbuch_paper_Jun-
_______ (March, 2004), ―Restructuring Our Electricity Networks to Promote
Decarbonization: Decentralization, Mass-Customization and Intermittent Renewables in
the 21st Century,‖ Tyndall Centre Working Paper No. 49;
www.tyndall.ac.uk/publications/working_papers/wp49.pdf _______ (June 2003) ―Is gas really cheapest?‖ Modern Power Systems, 17-19 _______ (February 2003) ―Determining the real cost: Why renewable power is more
cost-competitive than previously believed, Renewable Energy World,
_______ (May, 2003) ―The True Cost of Fossil-Fired Electricity in the EU: A
CAPM-based Approach,‖ Power Economics
_______ (February 2000) ―Getting It Right: The Real Cost Impacts of a Renewables
Portfolio Standard,‖ Public Utilities Fortnightly, February 15
? Shimon Awerbuch, 2005 7
– Sep 13, 2005 Draft
_______ (November 2000), ―Investing in Photovoltaics: Risk, Accounting and the
Value of New Technology,‖ Energy Policy, Special Issue, Vol. 28, No. 14
_______ (1995) ―New Economic Cost Perspectives For Valuing Solar Technologies,"
in, Karl W. Böer, (editor) Advances in Solar Energy: An Annual Review of Research
and Development, Vol. 10, Boulder: ASES
_______ (April 1995) ―Market-Based IRP: It‘s Easy!‖ Electricity Journal, Vol. 8, No. 3, 50-67
_______ (1993) "The Surprising Role of Risk and Discount Rates in Utility Integrated-
Resource Planning," The Electricity Journal, Vol. 6, No. 3, (April), 20-33.
_______ and M. Berger, (2003) Energy Security and Diversity in the EU: A Mean-
Variance Portfolio Approach, IEA Report Number EET/2003/03, Paris: February
http://library.iea.org/dbtw-wpd/textbase/papers/2003/port.pdf _______, Hyman, L. S. and Vesey, A. (1999), Unlocking the Benefits of Restructuring:
A Blueprint for Transmission, , Vienna, VA: PUR, Inc.
_______, J. Jansen and L. Beurskens, (2003), Portfolio-based Generation Planning:
Implications for Renewables and Energy Security, REEEP, British Foreign and Commonwealth Office, London, and United Nations Environment Programme, Paris,
_______ and Sauter, R. (2005), ―Exploiting the oil–GDP Effect to support Renewables
Deployment,‖ Energy Policy, articles in press, 21 June, www.sciencedirect _______, Stirling, A. C., Jansen J., and Beurskens, L.,  ―Portfolio and Diversity
Analysis of Energy Technologies Using Full-Spectrum Risk Measures,‖ in: D. Bodde,
K. Leggio and M. Taylor (Eds.) Understanding and Managing Business Risk in the
Electric Sector, Elsevier Topics in Global Energy Regulation, Finance and Policy
Brealey, R. and Myers, S. (2003) Principles of Corporate Finance, McGrawHill (any edition)
Dale, L. Milborrow, D. Slark R. & Strbac, G., (2004) Total Cost Estimates for Large-scale Wind Scenarios in UK, Energy Policy, Vol. 32, No. 17
DENA Grid Study (February, 2005), Planning of the Grid Integration of Wind Energy
in Germany Onshore and Offshore up to the Year 2020,
www.ewea.org/documents/Dena%20Grid%20Study_Summary_2005-03-16_dena.pdf DTI (2003) Economics Paper No. 4, Options for a Low-Carbon Future, June DTI (2004) DTI Energy Projections, ANNEX D: Electricity Generation by Fuel Type p.
74. http://www.dti.gov.uk/energy/inform/energy_projections/ep68_final.pdf IEA (2005) Projected Costs of Generating Electricity, Paris, Nuclear Energy Agency- International Energy Agency-OECD.
Reuters, (2005) ―Super Spike: Goldman Sachs predicts oil over $100,‖31 March ? Shimon Awerbuch, 2005 8