Distributed vs.

centralized electricity generation: are we witnessing

a change of paradigm?

An introduction to distributed generation

(HEC Paris)

Jeremi Martin

Under the supervision of Antoine Hyafil, Dean of the Energy Track at HEC Paris

May 2009

Abstract:

In this paper we study the reason for the current interest in distributed generation and the
challenges to be faced while increasing its share in the electricity generation mix. We decided
not put any restriction on the technologies used or plant size in the definition of distributed
generation and based our research on the definition putting the emphasis on the connection to
the distribution network and the proximity to the consumption point. We first describe the
centralized generation paradigm to show that its failure to provide answers for niche
generation markets combined with both electricity deregulation and the more stringent
environmental constraints paved the way an increasing share of distributed generation over
the past years. We then show that to be a credible alternative generation paradigm, distributed
generation will have to overcome significant technical, economic, regulatory and
environmental hurdles.

Keywords: distributed generation, decentralized energy production, cogeneration

Contact details: [email protected]

1
Executive summary

Under the current centralized generation paradigm, electricity is mainly

produced at large generation facilities, shipped though the transmission and
distribution grids to the end consumers. However, the recent quest for energy
efficiency and reliability and reduction of greenhouse gas emissions led to explore
possibilities to alter the current generation paradigm and increase its overall
performances. In this context, one of the best candidates to complement or even
replace the existing paradigm is distributed generation where electricity is produced
next to its point of use:

- historically, distributed generators have been able to act as a complement to

centralized generation i.e. they provided solutions to overcome the shortfalls
of the centralized generation paradigm;

- deregulation theoretically enabled distributed generators to enter the electricity

market through market price signals and fewer barrier to entry; and

- environmental concerns led regulators to promote efficient generation

technologies such as cogeneration. Cogeneration can be considered as
distributed as heat and steam cannot be easily shipped.

The main aim of this study is thus to better understand what hurdles currently
prevent distributed generation to play such a role. Based on recent research, we have
identified X different categories of barriers:

- technical issues: distribution networks will have to be reinforced and partly

redesigned to cope with new capacities. Besides, they will have to incorporate
both control and protection software and hardware to coordinate the
distributed generators and make.

2
- Price competitiveness: for such a move toward distributed generation to be
efficient in terms of both performance and price, distributed generation will
have to be used where it is more competitive than centralized generators i.e. at
congested areas where it is uneconomical to build a centralized plant or as
cogeneration facilities. This will also mean more research and development for
new technologies such as fuel cells in order to reduce the cost per kWh.

- regulatory barriers: significant work has to be undertaken to alter the

regulatory environment the distributed generators are facing: regulatory
hurdles still impede the spread of distributed generation as distribution
network operators have little incentive to give them access to the distribution
network while distributed generators are unable to cash in the positive impact
they have.

- environmental impact: distributed generation does not necessarily mean clean

generation. Indeed, diesel reciprocating engines often used as back-up
distributed generators tend to be the worst performers in terms of greenhouse
gas emissions. Distributed generation, to be a sustainable alternative paradigm,
will thus have to rely on the cleanest technologies or favour efficient uses
maximising energy efficiency and reducing emissions such as cogeneration.

3
 Content

Executive summary ........................................................................................................2

Introduction ....................................................................................................................5
I. Current state of the power market ...........................................................................8
A. The centralized paradigm ................................................................................8
1. Description and definition ...............................................................................8
2. The main drawbacks of the centralized paradigm ...........................................9
B. The main characteristics of distributed generation ........................................13
1. The main drivers behind the revival of distributed generation ......................13
2. Technologies used for distributed generation ................................................14
3. Current share of decentralized energy and prospective penetration rate .......18
II. The challenges to be faced while increasing the share of distributed generation.
20
A. Technical constraints: ....................................................................................20
B. Cost competitiveness: the economics of distributed generation ....................23
C. Regulatory barriers ........................................................................................26
D. Impact on climate change and global warming .............................................31
Conclusion and future research ....................................................................................35
References ....................................................................................................................37

4
 Introduction

In the current context of high fluctuation in energy prices, concerns over fossil
fuel depletion and increased awareness of greenhouse gas emission, the European
Union sees energy efficiency as a major challenge for the years to come. In its
communication, the Commission of the European communities (2006) estimates that
the European Union can save up to 20% of its energy consumption over the period
2007- 2020. In several countries such as the United Kingdom, a wide range of
possibilities are currently explored including distributed generation.

The Governments Energy Review Report of July 2006 highlighted the

challenges we face in addressing climate change and ensuring security of energy
supplies. A key part of responding to this challenge is to investigate to what extent DG
could complement, or in the longer term potentially offer an alternative to, a
centralised system. (Ofgem, 2007)

To better understand the implication of such a statement, special care has to be

given to the definition of distributed generation. As a relatively new field of research,
several expressions are still currently used such as decentralized generation,
dispersed generation, distributed energy resources etc. As show by Pepermans et
al. (2005), the definition varies significantly in terms of characteristics of the
generators mentioned. Dondi et al. (2002) define distributed generation as a generator
with small capacity close to its load that is not part of a centralized generation system.
Chambers (2001) puts a limitation on the maximum capacity of distributed generation
(30kW). There is however no consensus in the literature on the upper limit to be set:
this limit can range from 1MW to over 100MW (Ackermann et al., 2001). So as to
define distributed generation in a way that could encompass a wide variety of
technologies, capacities, the connection type and so forth, Ackermann et al. (2001)
devised a definition applicable to the vast majority of distributed generators.

Distributed generation is an electric power source connected directly to the

distribution network or on the customer site of the meter. (Ackermann et al., 2001).

5
 The key criteria in this definition are the connection to the distribution
network and the proximity to the end consumer. Using such a definition will for
instance rule out of the study large wind farms connected to the transmission network.
In a weaker form of the definition, we will include large combined heat and power
plants to the definition: large cogeneration facilities can be connected to the
transmission network but are conceptually close to distributed generation as they need
to be located in the vicinity of their heat consumers.

Put into an historical perspective, Ofgems aim to promote research in

distributed generation can come as a surprise as the past century has been driven by
centralization and increase in scale of power plants: though born decentralized,
electricity quickly moved to a centralized paradigm where electricity is produced,
transmitted and distributed to the end consumers. As early as 1930s centralized
generation was the leading form of generation in the US thanks to the advent of the
AC grid (Carley, 2009). The main goal of this study will thus be to understand why
we should reconsider an electricity generation system that was so easily dismissed in a
past and to identify the drivers of such a trend reversal i.e. the key advantages of
distributed generation that are at the outset of such a paradigm change. The first
objective of the study is to describe the current state of the power market where
centralized generation is dominant and distributed generation accounts for a relatively
small share of the total generation on average. The paper will focus on the main assets
of distributed generation, the technologies used and current and prospective
penetration rates. We will there review the different classifications of distributed
generators introduced in the literature to provide the reader with a consolidated view.
The second part will show that distributed generation currently lacks the inner
strength to become the dominant generation paradigm: increasing sustainably the
share of distributed generation will mean lifting several barriers that currently
impeded its spread. The hurdles are first of technical nature: the current network
architecture is able to handle a limited number of distributed generators but will have
to be revamped to accommodate a larger proportion. Second, given the current lack of
cost competitiveness of some technologies on a $/kWh basis, generators will have to
provide significant effort in research and development or favour use of distributed
generation where they can prove more competitive than centralized generation. Third,
the study will review the regulatory barriers to distributed generation expansion:

6
though deregulation had a positive impact on distributed generation, additional
regulation will be needed to ensure a balanced growth. The study will rely on research
results by Cossent et al. (2009) while conducting such a review. Last, based on recent
research by Strachan et Farell (2006) on the emission levels for different distributed
generation technologies, we will show that moving towards this new paradigm will
not ensure clean generation for all uses. Increasing the performance of distributed
generation with respect to emissions will necessarily mean a larger share of
cogeneration in the short run and further development of fuel cells in the medium run.

7
I. Current state of the power market

A. The centralized paradigm

1. Description and definition

Since the 1990s, electricity production has been driven towards generation
concentration and a higher degree of integration leading to the current centralized
electricity paradigm. This move was driven by several factors (US DOE, 2007):

- Economies of scale: the advent of steam turbines made it possible to increase

the size of the turbines while decreasing the marginal cost of electricity
production. The rapid spread of this technology led to a surge in the overall
plant capacities;

- The search for high energy efficiency: gains in efficiency were achieved
through larger facilities capable of handling higher pressures and temperatures
in steam used in electricity generation. At a certain point, the gains were
however offset by the increase in operating and maintenance costs as materials
were unable to sustain operation at high specification over the long run;

- Innovation in electricity transmission: the use of alternative current instead of

direct current permitted to transmit electricity over long distances with a
significant loss reduction;

- The search for reliability: so as to increase the reliability at the customers end,
large electricity production facilities were connected to the transmission
networks. Pooling resources helped reduce the reliance of each customer on a
particular generator as other generators were often able to compensate for the
loss;

8
 - Environmental constraints: the use of transmission networks made it possible
to relocate the generation facilities outside the city centres thus removing
pollution due to exhaust from coal fired plants;

- Regulation favouring larger generation facilities.

In the sectors layout resulting from this move towards concentration and
integration, electricity is generated, transported over long distances through the
transmission network and medium distances through the distribution network to be
finally used by the end customer. This can be summed up as follows:

Traditional electrical power system architectures reflect historical strategic

policy drivers for building large-scale, centralised, thermal- (hydro-carbon- and
nuclear-) based power stations providing bulk energy supplies to load centres
through integrated electricity transmission (high-voltage: 400, 275 and 132 kV) and
distribution (medium-, low-voltage: 33 kV, 11 kV, 3.3 kV and 440V) three-phase
systems. (Mc Donald, 2008).

Though dominant, centralized generation has always been operating along a

smaller distributed generation capacities that were never phased out of the market.
The persistence of the first historical form of energy generation whereby energy is
consumed near its generation point seems puzzling in the light of the properties of
centralized generation mentioned above. The significant size of distributed generation
in countries such as Denmark (detailed in section B) clearly implies that it is capable
of overcoming shortfalls of the centralized generation paradigm.

2. The main drawbacks of the centralized paradigm

Several studies were conducted to emphasize the main shortfalls of the

centralized generation paradigm and to explicit the motivation of the agents in
keeping distributed generation as a primary source of electricity or as a back up
generator (El-Khattam et Salama, 2004; Perpermans et al., 2005). The main drivers
listed in the literature are summarized below:

9
Transmission and distribution costs: transmission and distribution costs amount for up
to 30% of the cost of delivered electricity on average. The lowest cost is achieved by
industrial customers taking electricity at high to medium voltage and highest for small
customers taking electricity from the distribution network at low voltage (IEA, 2002).
The high price for transmission and distribution results mainly from losses made up
of:
- line losses: electricity is lost when flowing into the transmission and
distribution lines;
- unaccounted for electricity; and
- conversion losses when the characteristics of the power flow is changed to fit
the specifications of the network (e.g. changing the voltage while flowing
from the transmission network to the distribution network) (EIA, 2009).

The total amount of the losses is significant as shown in Table 1. In addition to

the cash cost, these electricity losses have an implicit cost in terms of greenhouse gas
emissions: fuel is consumed thus generating greenhouse gases to produce electricity
that is actually not used by the final consumer.

Table- 1. Transmission, distribution losses and unaccounted for

electricity in the U.S.

Date Net Generation -Bn kWh T&D losses and unaccounted for In %
1973 1864 165 8.9%
1975 1921 180 9.4%
1980 2290 216 9.4%
1985 2473 190 7.7%
1990 3038 203 6.7%
1995 3353 229 6.8%
1996 3444 231 6.7%
1997 3492 224 6.4%
1998 3620 221 6.1%
1999 3695 240 6.5%
2000 3803 244 6.4%
2001 3737 202 5.4%
2002 3858 248 6.4%
2003 3883 228 5.9%
2004 3971 266 6.7%
2005 4055 269 6.6%
2006 4065 266 6.5%
2007 4157 264 6.4%
2008 4115 241 5.9%

Source: Energy Information Administration, 2009

10
Rural electrification: in an integrated power system, rural electrification is challenging
for two reasons. As large capital expenditures are required to connect remote areas
due to the distance to be covered through overhead lines, connecting remote areas
with small consumption might prove uneconomical. This effect is amplified when
taking into account transmission and distribution losses because both tend to increase
with the distance covered. Rural electrification is thus costly. It often proves more
economical to rely on distributed generation in such cases (Carley, 2009). This has
often been the case for mountain areas or low density areas remote from the main
cities.

Investment in transmission and distribution networks: over the next 20 years,

significant investment will be required to upgrade the transmission and distribution
networks. The International Energy Agency (2003) estimated the total amount to be
invested in generation, transmission and distribution up to 2030 for the OECD
countries to stands between 3,000 and 3,500 billion dollars (base case predictions). In
order to cut these costs, distributed generation can be used as a way to bypass the
transmission and distribution networks. In its alternative scenario under this
scenario distributed generation and renewable energy are more heavily supported by
policy makers- the IEA forecasts the overall amount to be invested to be lower than
3,000 billion dollars (electricity generation investments remaining constant).

Energy efficiency: in the 1960s, the marginal gains in energy efficiency through size
increase and use of higher temperature and pressure started to diminish. Higher
temperatures and pressure resulted in high material wear and tear leading to lower
than expected operating life for steam turbines (Hirsch, 1989). In order to increase
energy efficiency without requiring to higher pressure, cogeneration systems have
been developed to reuse the waste steam in a neighbourhood heating system or
cooling system through district heating and/or cooling district. The total energy
efficiency achieved when combining both electricity and heat goes up to 90% (IPPC,
2007). Comparatively, the sole electricity generation hardly goes above 40%. The
main problem, however, is that steam and heat are even less easily transported than
electricity, thus justifying the use of distributed generation through production next to
the point of consumption.

11
Security and reliability: The persistence of distributed generation contributed to
energy security through two effects:

- Fuel diversity: as distributed generation technologies can accommodate a

larger range of fuel that centralized generation, distributed generation has been
used to diversify away from coal, fuel, natural gas and nuclear fuel (IEA,
2002). For instance, distributed generation has been used at landfills to collect
biogas and generate energy;

- Back up generation: the main use of distributed generation is for back up

capacities to prevent operational failures in case of network problems. Back-
up generators have been installed at critical location such as hospitals,
precincts etc.

Electricity deregulation and cost control device: in a deregulated electricity market,

the diminution of reserve margins or the failure of generators to supply the network
(due for example to unplanned outages etc) can lead to capacity shortfalls resulting in
high electricity prices to the consumers. In order to hedge against negative price
impacts, large electricity consumers have developed acquired distributed generation
capacities. Such a move was possible thanks to the increase in flexibility in the market
regulation following the deregulation including, among other, reducing barriers to
entry.

Environmental Impact: the environmental impact of the centralized energy system is

significant due to the heavy reliance on fuel, coal and to a lesser extent natural gas.
The electricity sector is responsible for of the NOx emissions, 1/3 of the CO2
emissions and 2/3 of the SO2 emissions in the United States (EPA, 2003). Distributed
generation has been used to mitigate the impact both in terms of emissions associated
with transmission and distribution losses, to increase efficiency through cogeneration
and distributed renewable energy.

As distributed generation has been able to overcome the aforementioned

shortfalls of the centralized generation paradigm, it kept on average a small share in

12
the overall generation mix. The following subsection will focus on the main features
of distributed generation and why it has been the source of an increased attention
recently.

B. The main characteristics of distributed generation

1. The main drivers behind the revival of distributed generation

As seen in the previous part, distributed generation has been historically used
in several ways to complement centralized generation. The reason behind the recent
revival of distributed generation is two-fold: the liberalisation of the electricity
markets and concerns over greenhouse gas emissions (Perpermans et al., 2005).
The electricity and gas deregulation process started in Europe following the
application of two directives (Directive 96/92/EC and 98/30/EC) aimed at providing a
free flow of gas and electricity across the continent. These directives and the
subsequent legislation created a new framework making it possible for distributed
generators to increase their share in the total electricity generation mix. The effect of
deregulation is two-fold (IEA, 2002):

- thanks to the reduction of barriers to entry and clearer prices signals,

distributed generators were able to move in niche markets and exploit failures
of centralized generation. Theses new applications took the form of standby
capacity generators, peaking generators (i.e. producing electricity only in case
of high price and consumption periods), generators improving reliability and
power capacities, generators providing a cheaper alternative to network use or
expansion, provision of grid support (i.e. provision of ancillary services
permitting better and safer operation of the network and/or shortening the
recovery time) (Pepermans et al., 2005).

- as distributed generators tend to be of smaller size and quicker to build,

they have been able to benefit from price premiums. Geographical and
operational flexibility made it possible to set up distributed generators in

13
 congested areas or use it only during consumption peaks. Besides, for small
excess demand, it is often uneconomical to build an additional centralized
generation plant whereas with lower CAPEX and capacities, distributed
generation might come in handy (IEA, 2002).

The second driver behind the rebirth of distributed generation is to be related

to environmental constraints. As shown by Pepermans et al. (2002), environmental
and economic constraints led to look for cleaner and more efficient use of energy.
Distributed generation has been able to achieve this target through:

- combined heat and power generation: using heat for central heating and
other applications makes it possible to reduce emissions and increase energy
efficiency to high levels. Cogeneration relies heavily on distributed generation
as heat transmission and storage is the source of significant energy losses;

- use of alternative fuel: distributed generation technologies make it possible

to accommodate a broad range of fuel. Typical applications involve collecting
waste gas from landfills and use it to generate electricity on site.

2. Technologies used for distributed generation

The variety of end-use is to be related to an even greater variety in

technologies. The range of technologies used for distributed generation and described
by the International Energy Agency (2002) includes:

- Reciprocating Engines: this technology uses compressed air and fuel. The
mixture is ignited by a spark to move a piston. The mechanical energy is then
converted into electrical energy. Reciprocating engines are a mature
technology and largely spread thanks to their low capital investment
requirement, fast start-up capabilities and high energy efficiency when
combined with heat recovery systems. Most reciprocating engines run either
on fuel or natural gas with an increasing number of engines running on biogas
produced from biomass and waste. On the rolling year June 2007- May 2008,

14
 most of the reciprocating engines ordered were used as back-up or stand-by
generators (45%1), the remaining being divided between peaking generators
(30%) and continuous generators (25%) (DGTW, 2008). Reciprocating
engines perform, however, poorly in terms of noise, maintenance and
emissions (IEA, 2002);

- Gas Turbines: gas turbines are widely used for electricity generation thanks to
the regulatory incentives induced to favour fuel diversification towards natural
gas and thanks to their low emission levels. Conversely to reciprocating
engines, gas turbines ordered over the period covered by the survey were
widely used as continuous generators (58%), 18% were used as standby
generators and 24% as peaking generators (DGTW, 2008). Gas turbines are
widely used in cogeneration;

- Microturbines: microturbines are built with the same characteristics than gas
turbines but with lower capacities and higher operating speed;

- Fuel cells: instead of converting mechanical energy into electrical energy, fuel
cells are built to convert chemical energy of a fuel into electricity. The fuel
used is generally natural gas or hydrogen. Fuel cells are a major field of
research and significant effort is put in reducing capital costs and increasing
efficiency which are the two main drawback of this technology;

- Renewable sources: renewable technologies have been used as a way to

produce distributed energy. Renewable sources range from photovoltaic
technologies, wind energy, thermal energy etc. These sources qualify as
distributed generation only if they meet the criteria of the definition which is
not always the case. Distributed generation is therefore clearly distinct from
renewable energy. For example, offshore wind farms do not qualify as
distributed generation.

1
The numbers are expressed in percent of unit orders and not in percent of total capacity

15
 The table below summarizes the main characteristics of the technology used as
presented by Perpermans and (2005).

16
Table- 2. DG technologies

Source: Pepermans et al., 2005

17
 3. Current share of decentralized energy and prospective
penetration rate

Estimating the worldwide share of distributed generation can lead to

significant divergence in results due to the differences in the definitions used. The
differences mentioned in the introduction can yield to significant adjustments in the
estimate of the total share of distributed generation. The inclusion or exclusion of
large cogeneration facilities can significantly affect the results. For instance, the total
share of distributed generation is of 2.5% in California if cogeneration capacities
larger than 20MW are excluded. If included, the share goes up to 17% of the total net
peak demand (Rawson and Sugar, 2007). The bone of contention here is whether this
large cogeneration capacities often connected to the transmission grid can be
considered as distributed generation. The impact on the result is even more significant
as the capacities of such facilities tend to be high.
A survey conducted by WADE in 2006 led to an estimated 25% share of
distributed generation (WADE, 2006). The definition used by WADE does not take
into account project size, the technology used or whether or not the facility is
connected to the distribution grid. It thus includes large cogeneration facilities. The
International Energy Agencys model (2003) estimates that distributed generation will
account for between 20 and 25% of additional capacities to be built up to 2030 in the
reference scenario and 30 to 35% in the alternative scenario. This will add up to
approximately 30 to 35% and 40 to 45% of power generation investments over the
period 2001 - 2030. The strong incentive to support distributed generation will be
driven, among others, by the size of investment in transmission network to be
avoided. It is estimated at $130 billion (IEA, 2003).

While focusing on the EU-25 countries, we see that distributed generation

accounts on average for around 10% of the total electricity production. As stressed out
in the chart below, the pattern significantly differs from one country to the other. In
Denmark, for instance, distributed generation accounts for more than 45% of the total
electricity produced (Cossent et al., 2009).

18
 Figure- 1. Distributed generation shares in the total electricity production
of the EU-25 countries (2004)

Cossent et al., 2009

The difference across countries can be explained by regulatory aspects,

political decisions e.g. France heavily promoted large nuclear power plants -, and
the way the electrical sector was historically built. In Denmark for instance, nuclear
power has never been an option for long term electricity generation due to public
concern over the risks and environmental impact related with this technology. Instead,
between 1970 and 2000, the government embarked in a vast program to promote
cogeneration, energy efficiency and renewable energy through wind power. The
implementation of such a program was done through a bottom up approach involving
a large number of small firms, municipalities and cooperatives working in close
cooperation (Lethonen and Nye, 2009).

19
II. The challenges to be faced while increasing the share of
distributed generation.

To sustain the forecast penetration rate, the architecture of the electricity

sectors needs to be altered. The current infrastructures were not originally built to
accommodate a large proportion of distributed generation. For the moment, only
necessary adjustments are undertaken in order to accommodate these new capacities.
The difficulties experienced by front-running countries such as Denmark give a
flavour of the problems to be faced as distributed generation is increasingly used.

For along time, western Denmark managed to increase DG connection

apparently with only minor changes in network control. []. It was not until the early
2000s that the reliability problems, created by the sudden increase in wind
generation, grew acutenotably the blackouts in eastern Denmark in 2003 were a
wake-up call.(Lethonen and Nye, 2009)

Over the long run, however, increasing significantly the share of distributed
generation will necessarily mean revamping the whole physical and regulatory
architecture of the electricity network and more precisely the distribution network.

A. Technical constraints:

The first difficulties to overcome are related to technical improvements

necessary to ensure high system reliability with distributed generation. The following
section gives an overview of the technical issues caused by distributed generation.
The classification and description are derived from a study by Pehnt and Schneider
(2006). The issues can be classified as follows:

Capacity: adding distributed generators at the distribution level can significantly

impact the amount of power to be handled by the equipment (cables, lines, and
transformers). In order to avoid overload problems, reinforcement work will have to

20
be undertaken. As shown by Pehnt and Schneider, the critical piece will often be the
transformers (converting medium voltage to low voltage or high voltage to medium
voltage): if power generated exceeds by far consumption, power will have to flow
back from the low voltage network to the medium voltage network or from the
medium to the high voltage network and be directed to other consumption areas. The
transformer will have to be able to handle this reverse flow i.e. being able to convert it
back and have specification to cope with potential oversupply. This is of major issue
at peak hours: at that time both continuous and peaking distributed generators will
operate to cash in the price premium. Production forecast from peaking distributed
generators is key while determining the specifications of the equipment, as capacities
will be added when the total power flow is already significant.

Voltage: distributed generators are often connected to low voltage networks. When
power is carried over long distance, voltage tends to drop due to resistance in cables.
As generators connected to the distribution network tend to increase the network
voltage, they may help keep the voltage within the specifications over the distance
and have a positive impact on the network. This positive impact is however strongly
dependent on the number of generators connected to the distribution network and their
concentration: above a certain threshold, adding another distributed generator might
negatively impact the network by increasing voltage above the specifications.

Protection: while using distributed generation, additional protection systems are

required to avoid internal faults, defective distributed network and islanding (Jenkins
et al., 2000). Islanding occurs when part of the network is still operating with the
distributed generators delivery electricity to customers while the rest of the network
has been disconnected. It can be useful to operate the network in such a way to ensure
steady supply of consumers with critical need for electricity or ensure that the
majority of the network is still operating while a section is under maintenance. The
main issue comes from undetected islanding as network operators might undertake
repair work and thus incurs significant risk for staff members.

Voltage and current transients: short term abnormal voltage or current oscillation may
occur as distributed generators are switched on or off. The result of these oscillations
can have a destabilizing effect on the network.

21
Transmission and distribution losses: one of the key advantages of distributed
generation is that it helps reduce transmission and distribution losses as distributed
generators are not connected to the transmission grid and some of them might even
choose to operate as captive plant for a client with thus limited use of the distribution
grid. Recent research has however shown that above a threshold (at very high
penetration rate and with generators concentrated in a specific area and all of them
feeding the distribution grid), the size of the transmission and distribution losses goes
up again (Mendez et al., 2002).

Ancillary Services: as of today all the ancillary services positively impacting the
quality of electricity delivered are provided by centralized generators. For example,
centralized generators are requested to keep capacities in excess of peak load to adjust
production in case of demand surge, to hold voltage control devices As the share of
distributed generation increases, distributed generators will have to provide a larger
share of these services.

In addition to the technical issues mentioned above, two fields of research will
have to be further investigated: active network and virtual power plants and
microgrids.
Historically, distribution networks have been less sophisticated than
transmission network as they were passively distributing energy from the transmission
networks to the customers. The coordination between the generators and the
adjustments in outputs were done directly at the transmission level. The integration of
distributed generation on a large scale will however require the distribution network
to be active in the sense that they will have to manage the flow coming from
centralized generation through the transmission lines, forecast the levels of output
from distributed generators (and especially peak generators), collect information,
devise start-up procedures in case of system failures, automation. This increased
level of complexity will require the development of management and control
procedures necessary to ensure quick and safe operation (McDonalds, 2008). The
change in network control and management can either rely on a centralized control
entity or several local controlling entities coordinated together. The latter architecture
was adopted in Denmark where the Cell Architecture project has been launched. The

22
aim of such a venture is to build a decentralized control system where the grid
management is handled by semi-autonomous entities. These entities will then be able
to operate jointly or as island in case of failure on part of the network thus ensuring a
minimum impact on a located system failure (Lethonen et Nye, 2009).
An extension of this idea is the concept of virtual power plant. A virtual
power plant is the coordination of several distributed generators in order to act as an
integrated plant (Feldmann, 2002; Jnig, 2002; Stephanblome et Bhner, 2002; Arndt
et Wagner, 2003). The plant is virtual as it is not in one place but made of the
aggregation of several units. The operation of such a plant required a strong
integration of information, communication and management systems (Pehnt, 2006).
One way of integrating small scale distributed generators is through a microgrid. As
of today, distributed generators are mainly integrated though medium voltage grids.
Significant research is however underway to facilitate the integration in low voltage
grids with local coordinating functions or microgrids (Costa et al., 2008).

B. Cost competitiveness: the economics of distributed generation

One key hurdle to overcome in a deregulated power market is the cost

competitiveness of distributed generation. This parameter varies, however, a lot from
one technology to the other mainly. One of the main reasons for such a difference is
the age of the technology and its current state of development. For example
reciprocating engines have been used for decades and are a mature technology while
fuel cells are still subject to significant research and development in order to become a
credible source of generation. According to Pehnt (2006) more than 5 years of
research and development will be needed for it to become a mature technology to be
adopted on a large scale. Table 3 below gives an indication of the capital costs,
operating and maintenance costs and fuel cost of the different technologies. These
parameters indicatives as they are highly sensitive to inputs (Strachan et Farell, 2006).
Data for fuel cells are also high sensitive to the progress in research and development
linked with this technology. All costs measures are in $2000. The following analysis
draws its results on key research results presented by Strachan and Farell (2006).

23
 Table- 3. Cost comparison between DG technologies

Unit size Capital cost Fixed O&M cost Variable O&M Fuel production costs
Efficiency (%HHV) (MWe) ($/kW) ($/kW-yr) (c/kWh) (c/kWh)
1999 2001
average average
Gas reciprocating engines 29 0.2 750 15 1 0.68 1.19
Diesel reciprocating engines 35 0.2 700 15 1 1.05 1.44
Micro-turbine 25 0.06 800 15 0.6 0.68 1.19
Fuel cell 38 0.1 3000 15 0.6 0.68 1.19
Gas turbine 29 10 480 15 0.55 0.68 1.19
CCGT (centralised generation) 50 200 550 15 0.55 0.68 1.19
CST (centralised generation) 33 500 1100 15 0.4 0.38 0.42

Source: Strachan et Farell, 2006

As shown in the table, for most of the costs (all the capital costs, fixed and
variable operating and maintenance costs), distributed generation technologies are less
or as competitive as combined cycle gas turbines. Coal steam turbines tend to have
higher capital costs but remain highly competitive due to their cheap fuel costs. On a
pure cost per kilowatt basis, distributed generation is clearly not the cheapest source
of generation.
The only way to tweak the picture and make distributed generation
competitive is to price in some of the key characteristics and positive externalities of
this generation technique. For instance, taking into account the ability of distributed
generation to produce both electricity and heat can modify the hierarchy. In this case
cogeneration will compete against production of centralized energy and costs related
to the use of an additional boiler to provide heating for a facility (Strachan et Farell,
2006). On the chart below, internal combustion engine (ICE) is a proxy for
reciprocating engines.

24
 Figure- 2. Indicative cost comparison between technologies

Source: Strachan et Farell, 2006

The HPR ratio used in the chart above is defined as:

Energy produced (or consumed) as heat

HPR =
Energy produced (or consumed) as electricity

A heat to power of 0 is thus related to a case where electricity only is produced

while a heat to power ratio of 2 is achieved under a situation whereby the facility
needs twice more heat than electricity. An increase in heat to power ratio has a
positive effect on the cost of distributed generation until the maximum steam
generation output of the technology is reached. The curves converge as when the HPR
is high: at this point cogeneration is no longer sufficient to produce heat and
additional boilers should be added. At high HPR ratios, the marginal cost is quasi only
influenced by the marginal cost of an extra boiler and no longer by the marginal cost
of the distributed generation technology (Strachan et Farell, 2006).

Cost competitiveness of distributed generation today is thus heavily impacted

by the capacity of the regulation to price its impact on the electricity network and on
its ability to provide specific services to the end consumer such as heat generation or
ancillary services. The diffusion rate of distributed generation will thus be driven by
the ability of the regulator to lift the regulatory barriers. This issue will be of major

25
importance as it will strike the balance between a market driven diffusion model
whereby the distributed generators will be able to increase their return on investment
through prices (incorporating factors such as avoidance of use of the grid, investment
deferral, emission abatements etc) and a subsidized model with the state or electricity
regulator imposing higher tariffs for distributed generators when the price signal fails.

C. Regulatory barriers

In addition to technical and cost issues mentioned above, distributed

generations share is still negatively impacted by the lack or the inappropriateness of
regulation. In a recent study, Cossent et al. (2009) made a review of European
regulation and analysed the key factors that might impede the spread of distributed
generation in Europe. The classification and main issues mentioned below are derived
from this study. The main issues identified are:

Network tariffs: when getting connected to the electricity for the distribution network,
distributed generators have to pay to the distribution network operator charges as a
remuneration of work undertaken. The charges are divided between connection
charges for the physical connection to the network (paid at once) and use-of-system
charges (paid on a recurring basis).
Connection charges can be divided between deep connection charges and
shallow connection charges. Deep and shallow connection charges can be defined as
follows:

Under deep connection charges, DG pays for all the cost of connection,
including upstream network reinforcements. On the other hand, under shallow
connection charges DG pays only the direct costs of connection (Cossent et al.,
2009).

The choice between deep and shallow connection charges is bound to have a
major impact on the penetration rate of distributed generation. Deep connection
charges will be detrimental for small scale distributed generators and to some extend
peaking distributed generators: the investment needed for connection will

26
significantly reduce the net present value of the investment and can to some extent
make it become negative. The differences in regulation across European countries
thus make investment more or less attractive in member state countries when taking
into account this criterion (Table 4). The term shallow-ish refers to a situation where
the distributed generation pays for direct connection costs and part of the
reinforcement of the system proportional to its system use.
Use-of-system charges are generally not a major issue for distributed
generators: the regulation is generally favourable to them as they are not required to
pay it. This feature is bound to change if distributed generation accounts for a large
share of total generation.

27
Table- 4. Review of distributed generation regulation

Source: Cossent et al., 2009

28
Planning: one of the main benefits of distributed generation is to defer costly
investments in distribution networks (upgrade or capacity increase) by producing
electricity where it is most needed. Though cost effective, these economies are often
not realised due to the structure of the revenues of the distribution network operators.
As regulated natural monopolies, distribution network operators are often
remunerated on a cost plus of rate of return basis with adjustment for reaching
performance tests. This structure gives them little interest in favouring distributed
generation as they do not directly benefit from this improvement: operators will
choose to invest in the costly solution that gives them a safe income (i.e. the network
extension or upgrade with a guaranteed rate of return) instead of a less costly solution
with no gain. This is even more pronounced as since the electricity market
deregulation, transmission network operators cannot hold generating capacities. In
some countries, regulation has been adapted to take into account this potential cost
reduction.

Incremental distribution costs caused by distributed generation: though in the long run
distributed generation defers investment in the network, reinforcement work has to be
undertaken to accommodate this new form of generation. This additional distribution
costs (incremental CAPEX and OPEX) caused by distributed generation is seldom
accounted for in the current compensation of network operators. Operators are thus
less inclined to favour this option.

Energy losses: the treatment of energy losses varies greatly across countries. With a
low penetration rate and low concentration, distributed generation has a positive
impact on these losses. Regulation on this specific point affects the profitability of the
distributed generators. In countries such as Italy, the network operator pay for the loss
avoidance to the distributed generators connected at the transmission level. In France
however, the operator has incentive to encourage distributed generators to enter the
market as he pays for the losses through the purchase of electricity to a centralized
generator (Cossent et al. 2009). The network operator thus has a financial gain in
letting distributed generators enter the market and reduce this amount of losses. This
is however less advantageous to the centralized generators. The adoption of such
solution will thus be significantly affected by the relationship between centralized
generators and network operators. When both, though legally independent from one to

29
the other, are owned by the same entity, the operator might not be willing to favour
the distributed generators.

Ancillary services: distributed generators can help improve the quality of services
provided through voltage control (connecting a distributed generator to a low voltage
network makes it possible to reduce the drop in voltage over the distance), providing
additional peaking power capacities . These potential services to be rendered are
not a source of revenues for distributed generators under the current regulation.
Regulatory change might be phased in this impact. One of the key hindrances here,
besides regulation, is the lack of long term historical data to assess the overall impact
of these generators.

Incentives for innovation: integrating distributed generators into the distribution

network will trigger a major change in the management and control process. The
replacement of traditional grids by smart grids is a prerequisite for a larger share of
distributed generation. The results of the regulatory incentives given for innovation
are mixed.

Unbundling activities: The operators independence with respect to centralized

generators is key for distributed generation. Electricity deregulation ensures legal
independence of the two entities. Legal independence is, however, not sufficient to
ensure that network operators will not act in favour of centralized generators. The
problem is even more acute when ownership structures are the same (Praetorius,
2006). Studies even report unfair discrimination towards distributed generators (Jrss
et al. 2003). This problem is of major importance as distributed generators are often in
a critical phase of their investment process when encountering such difficulties. Due
to their relative small size, unjustified delays or procedures might significantly affect
their financial strength.

In addition to these regulatory barriers, Ofgem (2007) reported in its review of

distributed generation additional regulatory burdens in the form of:

30
 - licensing requirements that were devised for large centralized generators:
the requirements are often irrelevant for small scale generators or come at a
high cost;

- difficulties in getting permissions and delays;

- issues with exported electricity: one key complement of distributed

generators is to be able to supply the grid when needed or when consumption
at the nearest point is too low. The lack of visibility on the tariffs or the low
tariffs for the electricity fed back to the grid has been a major source of
concern for distributed generators. This problem has a major impact on the
economic profitability of the project.

Last, network operators lack incentives to take into account heat in the case of
distributed cogeneration. Operators will favour project having a positive impact on the
system stability regardless of the increased efficiency to be achieved through
production of electricity and heat (Woodman et Baker, 2008). This is unfortunate as
cogeneration is one of the key assets of distributed generation.

D. Impact on climate change and global warming

As previously mentioned, distributed generation does not necessarily mean

clean generation. The aim of this section is to better understand the environmental
impact of the main technologies used in distributed generation.
Strachan and Farell (2006) analysed the performance with respect to emissions
of distributed generation operating either for the sole production of electricity or as
cogeneration units. Renewable energies were left off the sample as for those
technologies the main concern is less emissions than cost per kWh. Table 5 gives the
emissions comparison between gas and diesel reciprocating engines, microturbines,
fuel cells, combined cycle gas turbines and coal steam turbines. The emissions
produced by a boiler have been added to make it possible to compare the emission of
centralized generation with an extra boiler used on site for heat and distributed
cogeneration.

31
 Table- 5. Emission factor for distributed and centralized generation

CO2 (g/kWh) SO2 (g/kWh) NoX (g/kWh) CO (g/kWh) PM10 (g/kWh) HC (g/kWh)
Gas reciprocating engine 625 0.032 0.5 1.8 0.014 0.54
Diesel reciprocating engine 695 1.25 2.13 2.8 0.36 1.65
Micro-turbine 725 0.037 0.2 0.47 0.041 0.14
Fuel cell 477 0.024 0.015 0 0 0
Gas turbine 625 0.032 0.29 0.42 0.041 0.42
CCGT 363 0.019 0.195 0.07 0.041 0.05
CST 965 5.64 1.7 0.07 0.136 0.05
Gas-fired boiler 201 0.01 0.12 0.12 0.01 0.014

Source: Strachan and Farell, 2006

As already mentioned above distributed generators are not best in class when
it comes to emissions. Diesel reciprocating engines are the worst emitters in terms of
nitrogen oxides (NOx). Combined cycle gas turbines tend to be the best performers in
terms of carbon dioxide (CO2), sulphur dioxide (SO2) while fuel cells are the lowest
emitters when it comes to NOx, CO, particulate matters (PM10) and hydrocarbons
(HC).
Starchan and Farell then complete the analysis taking into account the
possibility to use the heat produced by distributed generators for heating a facility.
The charts below plot the emission against the heat to power ratio. Heat to power
ratios tend to vary significantly across the regions and seasons with for instance a heat
to power ratio for aggregated seasonal demand of 3 for New York in Winter and 0.82
for Florida in Summer as shown in the study. The optimal technology to be used here
is thus a function of costs, emissions and heat to power ratio.

32
Figure- 3. Emissions comparison between centralized and
distributed generation when both electricity and heat are
needed

Source: Strachan and Farell, 2006

33
 As shown in the charts above, the results are somewhat mixed. The highest
emitters tend in general to be the diesel reciprocating engines and the coal steam
turbines. This fact has been known for a long time and is at the heart of regulation
implemented to reduce the emissions at CST plants while imposing more stringent
conditions on the use of diesel reciprocating engines as back-up generators. The
lowest emitting technology is the fuel cell but at high HPR as to little heat is produced
additional boilers have to be used. As shown in the cost comparison figure, this
technology is still crippled by high costs.

Distributed generation is not always the best performer in terms emissions.

Though some technologies such as fuel cells seems promising for future application,
the absence of implementation background and the costs of this technology are still
hindering its diffusion on a large scale. To be used in the cleanest way possible,
distributed generation will thus have to use the less emitting technologies and favour
cogeneration.

34
 Conclusion and future research

The recent attention paid to distributed generation is not surprising in many

ways. The flaws of the centralized generation paradigm led to look for a complement
generation technique, a role that was endorsed to the extent possible by the distributed
generators before the energy deregulation. The drastic change in the environment of
the electricity generators in the 1990s laid the foundation for an increasing use of
distributed generation. As shown by Pepermans et al. (2005), energy deregulation and
concerns over human impact on climate change created a context in which the revival
of distributed generation became possible: clearer price signal and fewer barriers to
entry made it possible for small scale generators to enter niche markets while
environment concerns positively impacted efficient generation methods such as
cogeneration. The renewed interest for the oldest generation paradigm (at the
beginning of the electricity industry all plants were distributed), made us wonder if it
can account for a larger share in the total generation. Forecast penetration rates are
high but rely on strong assumptions on the future state of the electricity market
(regulation, government backing of a distributed generation policy etc). Besides,
though front-running countries such as Denmark made it clear that it can account for a
large fraction of generation, we were still wondering if such an example was not in
the end an exception: the size of distributed generation increased to such levels mainly
through cogeneration due to the high demand for heat related to the low temperatures
in winter.
The in-depth study of the electricity market structure and of the key
characteristics of distributed generation demonstrated distributed generation can
consider becoming an alternative paradigm provided it can overcome significant
obstacles. First technical issues will need to be solved: better understanding of the
impact of a high number of distributed generators need to be developed and network
management procedures will have to be overhauled to fully exploit its implicit
benefits. From an economic perspective, distributed generators will have to go on
more research and development to lower the cost per kWh while specialising in end-
use where they can prove more competitive than centralized generators. Besides, the
growth of distributed generation will be closely linked to the capacity of the
generators to monetize their positive impact on the overall electricity sector. This will

35
only be possible if the sector and prices regulation integrates this effect through an
adjustment in the price of electricity paid to the generators. The new legislation to be
implemented will thus be crucial in defining a diffusion of model of generation (a
market model or a subsidized model). As the revival was partly driven by
environmental concerns, effort on total emissions from distributed generators will
have to be the object of more scrutiny. Distributed generation is not the cleanest and
the most efficient source of generation, at least for non renewable generators. The best
solution to increase its performance is to use if for cogeneration which implies a
greater integration on electricity heat and cooling network. Though a widely used
technology, cogeneration still faces difficulties as there is in general no link between
regulators for electricity and heat (Woodman et Baker, 2008). Further research will
thus investigate the change needed to ensure better use of cogeneration. This will also
require questioning the impact on a third driver currently reshaping the energy sector,
the search for supply security: most of the cogeneration facilities are currently running
on natural gas.

36
 References

Ackermann, T., Andersson, G., S.oder, L., (2001). Distributed generation: a

definition. Electric Power Systems Research 57, pp. 195204.

Arndt, U., Wagner, U., (2003). Energiewirtschafltiche Auswirkungen eines

Virtuellen Brennstoffzellen-Kraftwerks. VDI-Berichte.

Carley, S., (2009). Distributed generation: an empirical analysis of primary

motivators. Energy Policy, 37, pp. 1648 1659.

Chambers, A., (2001). Distributed generation: a nontechnical guide. PennWell,

Tulsa, OK, pp. 283.

Cossent,R., Gomez T., Frias, P., (2009). Towards a future with large penetration of
distributed generation: Is the current regulation of electricity distribution ready?
Regulatory recommendations under a European perspective. Energy Policy, 37,pp.
1145-1155.

Costa, M.P., Matos, M.A., Pecas Lopes, J.A., (2008). Regulation of microgeneration
and microgrids. Energy Policy, 36, pp. 3893 - 3904.

Dondi, P., Bayoumi, D., Haederli, C., Julian, D., Suter, M., (2002). Network
integration of distributed power generation. Journal of Power Sources 106, pp. 19.

DGTW., (2008). 2008 32nd Power Generation Order Survey, Diesel and Gas Turbine
Worldwide (http://www.dieselgasturbine.com/pdf/power_2008.pdf#zoom=100).

EIA (2009). Monthly energy review March 2009. Energy Information

Administration., (http://www.eia.doe.gov/mer).

El-Kattam,W., Salama, M.M.A, (2004). Distributed Generation Technologies:

definitions and benefits. Electric Power Research, 71, pp. 119-128.

37
European Communities, 2006. Communication from the Commission to the European
Council and the European Parliament. COM (2006) 545 final. Action Plan for
Energy Efficiency: Realizing the Potential.
(http://ec.europa.eu/energy/action_plan_energy_efficiency/doc/com_2006_0545_en.p
df).

Feldmann, W., (2002). Dezentrale Energieversorgung-zuknftige Entwicklungen,

technische Anforderungen. Conference; Energie Innovativ 2002. VDI Verlag,
Dsserldorf.

Hirsh, R.F., (1989). Technology and Transformation in the American Electric Utility
Industry, Cambridge University Press: New York, New York.

International Energy Agency, (2002). Distributed Generation in a liberalized energy

market. Jouve: France.

International Energy Agency, (2003). Distributed Generation and Renewables

Outlook 2030, (http://www.iea.org/textbase/work/2004/distgen/Birol.pdf).

Intergovernmental Panel on Climate Change (IPCC), (2007). Climate Change 2007

Mitigation of Climate Change: Working Group III contribution to the Fourth
Assessment Report of the IPCC (Climate Change 2007), IPCC, Cambridge University
Press.

Jnig, C., (2002). Perspektive. Lokal Energie Geschftsbericht 2001. Stadtwerke

Unna GmbH, Unna.

Jenkins, N., Allan, R., Kirschen, D., Strbac, G.,(2000). Embedded Generation.
Institution of Electrical Engineers (IEE), London.

Jrss, W., Jorgensen, B.H., Lffler, P., Morthorst, P.E., Uyterlinde, M., Sambeek, E.,
Wehnert, T., (2003). Decentralized Power Generation in the Liberalized EU Energy
Markets. Springer: Berlin Heidelberg New York.

38
Lehtonen, M., Nye, S., (2009). History of electricity network control and distributed
generation in the UK and Western Denmark. Energy Policy,
doi:10.1016/j.enpol.2009.01.026.

McDonald, J., (2008). Adaptive intelligent power systems: active distribution

networks. Energy Policy 36, pp. 43464351.

Mendez, V.H., Rivier, J., de la Fuente, J.I, Gomez, T., Arceluz, J., Marin, J., (2002).
Impact of Distributed Generation on Distribution >etwork. Universidad Pontificia
Comillas, Madrid.

OFGEM, 2007. Review of Distributed Generation. Office of Gas and Electricity

Markets (www.dti.gov.uk/energy/whitepaper).

Pehnt, M., (2006). Micro Cogeneration Technology, in Pehnt, M., Cames, M.,
Fischer, C., Praetorius, B., Shneider, L., Schumacher, K., Voss, J.P. Micro
cogeneration towards decentralized energy systems, Berlin: Springer, pp. 197-218.

Pehnt, M., Schneider, L., (2006). Embedding Micro Cogeneration in the Energy
Supply System, in Pehnt, M., Cames, M., Fischer, C., Praetorius, B., Shneider, L.,
Schumacher, K., Voss, J.P. Micro cogeneration towards decentralized energy
systems, Berlin: Springer, pp. 197-218.

Pepermans, G., Driesen, J., Haeseldonckx, D., Belmans, R., Dhaeseleer, W., (2005).
Distributed Generation: definition, benefits and issues. Energy Policy, 33, pp. 787-
798.

Praetorius, B., (2006). Micro Cogeneration Setting of an Emerging Market, in

Pehnt, M., Cames, M., Fischer, C., Praetorius, B., Shneider, L., Schumacher, K.,
Voss, J.P. Micro cogeneration towards decentralized energy systems, Berlin:
Springer, pp. 145-170.

39
Rawson, M.,Sugar J., (2007). Distributed Generation and Cogeneration policy
roadmap for California. California Energy Commission
(http://www.energy.ca.gov/2007publications/CEC-500-2007-021/CEC-500-2007-
021.PDF).

Stephanblome, T., Bhner, V., (2002). Virtuelles Kraftwerk: Energiewirtschafliche

Voraussetzungen & Leittechnick-Software. Conference: Energie Innovativ 2002.
VDI Verlag, Dsseldorf.

Strachan, N., Farell, A., (2006). Emissions from distributed vs. Centralized
generation: the importance of system performance. Energy Policy 34, pp. 2677-
2689.

US DOE, (2007). The Potential Benefits of Distributed Generation and rate-related

issues that may impede their expansion. A Study pursuant to Section 1817 of the
Energy Policy Act of 2005.US Department of Energy.

US EPA., (2003). Inventory of US Greenhouse Gas Emissions and Sinks: 19902001.

EPA 430-R-03-004. US Environmental Protection Agency, Washington, DC.

WADE., (2006). World Survey of Decentralized Energy, 2006. World Alliance for
Decentralized Energy.
(http://www.localpower.org/nar_publications.html).

Woodman, B., Baker, P., (2008). Regulatory framework for decentralised energy.
Energy Policy 36 , pp. 45274531.

40