You are on page 1of 6

CICED2008

Technical Technical Session 5

Development Planning of Distribution

S5-22 FP1762

Advanced Planning Methods Maintaining Reliability and Power Quality in


Networks with Increased Distributed Generation
G. Mehlmann', M. Bohm2, T. Ken J , J. Jiger4
Erlanger Stadwerke AG (public utility company ofErlangen), Erlangen 91058, Germany
Institute of Electrical Power Systems, University of Erlangen, Erlangen 91058, Germany
E-MAIL: gert.mehlmann@estw.de(ormehlmann@eev.eeLuni-erlangen.de).michael.boehm@estw.de.
keil@eev.eeLuni-erlangen.de, j aeger@eev.eeLuni-erlangen. de
1,2

3,4

Abstract:
Distributed generation (DG) is finding its way into the
worldwide distribution networks very rapidly. As the existing
networks were not created for this new structure of electricity
production, the preservation of a reliable network operation and
a high level of power quality is a big challenge for all network
operators. In a study conducted by the University of Erlangen
and the Erlanger Stadtwerke AG (public utility company of
Erlangen), this problem was investigated using an existing
distribution network. In this paper the study and its results are
presented. Different future scenarios of network configurations
with increased DG will be shown. On the basis of calculations for
each scenario regarding load and voltage profile, short-circuit
currents and (n-I)-aspects the effect of DG on the distribution
network will be pointed out. Using these results methods will be
presented that guarantee a secure operation in those scenarios.
In addition, necessary investment costs are registered in this
paper.

Keywords:
Distributed generation; co-generator; PV plant; greenfield
network; future scenario

1.

Introduction

The declining of fossil fuels and increasing


environmental consciousness causes an increasingly
distributed power generation worldwide. Thus DG is
supported by fmancial incentives. Consequently a strongly
increasing number of third parties as private entities become
energy provider. Commonly their infeed connection is at low
voltage level [7]. This new pattern of distribution causes new
network conditions [6], [8]. Fluctuating of infeed,
bidirectional powerflow, difficult protection layout, changing
loadings of equipment and voltage fluctuations are only some
of the challenges network operators will have to take on [2].
The low voltage network is of particular interest, due to two
basic principles: As its expansion is historically it is not
comparably reliable like the middle or high voltage network

where the (n-l )-principle is strictly implemented. In addition,


the infeed of many small generators located across a large
area is technologically more difficult to handle than the infeed
of single controllable plants with high capacity.
Using the planning software PSS Sincal an existing low
voltage network of an urban area was modeled. All following
analysis has been done using this model. One fundamental
goal was to do all investigations without stochastic
valuations. Therefore, creating the scenarios calculations had
to be done for each local unit in the area in consideration of
their economic efficiency and their technical use concerning
DG [3].

2.

Network

The network is a meshed system with more than 2200


measuring points (clients) in an urban area. Because of its
structure it is attractive for combined heat and power units
(co-generators) and solar energy plants. It extends to 0,4 km2
and is feeded by six transformers, four 630 kVA transformers
and two 500 kVA transformers. Therefore, the nominal
apparent output is 3520 kVA. Mains supplies are connected
upstream of the transformers. Figure 1 shows the used
symbols; Figure 2 the current network in PSS Sincal [5].

mains
supply

1~ ~ ~

load

PV plant

cotransformer cable choke


generator

Figure 1: Used Symbols in Figure 2

CICED2008

Technical Technical Session 5

Development Planning of Distribution

highest equity return but also the lowest profit. For all types
of PV plants the profit is higher for a higher equity ratio. But
the equity return is only growing if the profit is rising faster
than the equity ratio. For the amorphous PV plants the equity
return is often declining because of the low increase of the
profit with the equity ratio. However, only amorphous plants
are making profit with a high equity return in the range of
very small peak power and strong outside fmance.
In Germany the best way to configure a heat controlled cogenerator is for own consumption. Only the excess capacity
should be used to supply other consumers with current.
Specifically, the electricity tariff namely exceeds the subsidy
for conventional and EEG co-generators. While the baseload
price is rising, the subsidy depending on EEG is declining and
will be overrun by the baseload price. As a result of these
facts, the profitability of a co-generator depends on the
consumption of the corresponding building. EEG cogenerators have high investment costs and high subsidies;
conventional generators lower investment costs but also lower
subsidies. That means, the higher the consumption the more
lucrative the co-generator, especially the conventional ones.
However if the consumption is lower than the production, the
co-generators, especially the conventional generators, are not
working efficient. Therefore, big buildings of modem
standards are best suited for co-generators, as they have a
high consumption of current but only a low heatconsumption. Another important aspect is the development of
the electricity tariff compared with the costs of resources.
Only if the electricity tariff is rising faster than the costs of
the resources, co-generators for private use will have a broadbased future. Overall conventional co-generators are more
profitable than EEG co-generators.
Depending on their profitability the plants were
separated into classes. With these classes as well as with cost
trends and the urban development it was possible to develop
the future scenarios.
The following four scenarios will be presented in succession.

3.

Future Scenarios

Through the use of software coded for this study it was


analysed which parcel in the investigated area is suitable for
DG. In this way, the optimal plant was identified and
dimensioned for each place. According to each type of plant
the infeed is varying. Calculations for polycrystalline,
amorphous, monocrystalline and improved monocrystalline
PV plants were done. Additionally, calculations for the cogenerators were done. Two groups of co-generators have been
analysed: co-generators reimbursed by baseload prices that
work with conventional commodities like fuel or gas and
EEG (renewable energies act) co-generators that work with
regenerative energies. All co-generators are dimensioned as
heat controlled peak load plants for private use.
The economic examination of the supplies resulted in
some interesting facts that will be pointed out below.
Monocrystalline PV plants are inapplicable for
commercial use. Also the market share of amorphous PV
plants will increase. By trend, amorphous PV plants have the

Scenario 1 - Maximum infeed


In this scenario the consequences of a very high but only
theoretically possible increase ofDG are analysed.
Scenario 2 - Only PV plants
The intention of this scenario is to research, if it is possible to
integrate a very high number ofPV plants into city.
Scenario 3 - Only co-generators
Comparable to the second scenario the impacts of a very high
number of co-generators on the network are analysed.
Scenario 4 - Realistic scenario
The scenarios above are worst case scenarios that are
necessary to point out boundaries. This future scenario is
based on a realistic development of the investigated area.
All calculations in these scenarios were done for the
existing meshed network and a deduced radial system. At first
the results for the mesh system are presented.

CICED2008

Technical Technical Session 5

Development Planning of Distribution

3.1. Scenario 1- Maximum infeed


In this scenario a plant is constructed at every location
where economical loss-free operation is possible. Economic
objectives, such as a high equity return or a fast amortisation
are no criterion for the habitat. Already installed cogenerators are replaced with co-generators of higher capacity.
All in all the peak power of the PV plants is near 2,3 MW.
The nominal power of all co-generators is higher than 3,2
MW. As the PV Plants never reach their peak power but the
co-generators reach their nominal power over time, the infeed
of the co-generators is much higher.
The high infeed of distributed energy has detrimental
effects on the network.
The short circuit power and the short circuit currents are
rising rapidly. As a result of this, stresses in case of a fault are
much higher. Especially the co-generators that are mostly
based on rotating generators directly connected to the
distribution network boost the short circuit current. PV plants
and co-generators working with converters have only a low
effect on short circuit currents. Examples for those generators
are co-generators with stirling engines or fuel cells.
During normal operation the load of the transformers is
temporarily higher than 120 % of its nominal power. During
(n-Ij-operation the load is reaching 180 % for several hours a
day. In addition, the load current is now because of the DG
often higher than the allowed nominal current of the fuses,
because of the DG. Therefore, the distance between the
operating current and a fault current is too small, resulting in
a reliable protection is no longer being possible. A new layout
for the whole protection system would be necessary.
Due to strongly varying and non controllable infeeds
and loads, strong voltage fluctuations with local and temporal
sinks and maxima appear. At times with a high distributed
infeed and low request of power many node voltages exceed
the nominal voltage by even more than + I0 %. Therefore, the
tolerance for voltage quality according to DIN EN 50160
cannot be kept [1].
Load curve analyses demonstrate that there is a high
feedback of power on cold days (winter days) because of the
co-generators. Contrary to expectations even on days with an
eminent infeed of solar energy (summer days) feedback of
power is only during a short time and not very high. Figure 3
shows a simulated load curve for a winter day and Figure 4
for a cloudless summer day.
Furthermore a correlation between load density and
infeed density, illustrated in Figure 5, can be pointed out.
Overall the network has reached a critical point because
ofDG and without counteracting, operation is not sustainable.

3.2. Scenario 2 and 3


The results of scenario 2 and 3 show, that it is possible
to integrate even a very high number of PV plants into the
network without any difficulty. The roof areas in particular
are a natural limitation for the installable peak power. A high
number of co-generators results unavoidable in an insecure
mains operation. This is especially the case in areas with a
strong heat demand. A comparison of these scenarios with

scenario 1 shows, that the impacts of PV plants and cogenerators on the network are almost decoupled. This results
from the big differences of the load profiles by day and year.
0:00
0,0 .,..--

12:00
- -

24:00
t[h]

-0,5

-1,0

-1,5

-2,0
P[MW ]

Figure 3: Load curve (winter day)


P[MW ]

::I

1,0 ~
0,8
0,6

0,4
0,2

---~------"'c___1''---~----~trh]
0:00
24:00
-0,2
0,0

+-:

Figure 4: Load curve (summer day)

CICED2008

Technical Technical Session 5

Development Planning of Distribution

3.3. Scenario 4

This scenario represents a realistic development status at


about 2025. The installed peak power of the PV plants is near
1,6 MW and the nominal power of the co-generator near 1,4
MW. The increase of the short circuit power is now much
lower than in scenario 1 and the load of the transformers
during nominal operation remains under 70 %. This is even
lower as in the current network. In addition, (n-l )-operation is
possible without any difficulty and voltage tolerances
according to DIN EN 50169 can be kept [1]. There is also no
feedback of power. All in all DG tends to result in a load
rejection of the electrical equipment and cables in the near
future. But there are still problems with the load current
exceeding the allowed nominal current of the fuses. In Figure
6 and Figure 7 two load curves are exemplarily presented, one
for a winter day and one for a summer day.
P[MW]

1,2

1,0

network was converted by fitting cut-off points. Afterwards


all calculations for the scenarios were repeated.
The analyses in the radial distribution system
demonstrate that it is much more difficult to keep voltage
tolerances. The loads of the electrical equipment and cables
have increased as well as the electrical losses. The short
circuit power is much lower than in the meshed network. All
in all DO causes more problems in a radial system than in a
meshed system, albeit there are some advantages in the event
of a short circuit in the network.
4.

Planning objectives

Different planning objectives can be chosen to ensure a


reliable network operation and high level of power quality.
On the one hand it is possible to take some single measures,
like outlined in Figure 8. On the other hand it is possible to
design a new greenfield distribution system for DG. But even
in a network with many local supply systems, it is often
sufficient to do some selective touching up to guarantee a
supply of high quality. Therefore, investments for a new
network will not be necessary.

0.8

1'\

0,6

0,4
\

0,2 \

\
\ .....1''

0,0
0:00

;'

1,,--

, \1
I

\-.

V\-,/

~j

\.--J"/

12:00

Figure 6: Load curve (winter day)

s ho rt c irc uit
li mitat ion

s upply of

capacitive reactive

P[MW)

1,4

bette r vo ltage
q ua lity

1.2

1,0
0,8

technical
meas ures of t he
pup lic ut ility
co mpa nies

s ma lle r
d ime ns io nig o f
t he di stribu ted
s up plies

power

lower load of

the electrica l

0,6

equipment and
cables

higher
high costs for the
general public

0,4

lower

bene~::~roh~~~a~~~sts for the ~o:~~~:~:


and cables

0,2
0,0 1------~ ------~----- :-.- I [hJ
0:00
24:00
12:00

Figure 7: Load curve (summer day)


In contrast to scenario 1, the operation of the network
with DG is widely unproblematic, in many points even better
than in the current state.
3.4. Results for a radial distribution system

As more and more public utilities operate with radial


distribution systems the effects of DG on this network
configuration has to be analysed [4]. Therefore the existing

Figure 8: Measures to adapt a distribution system for DG


Planning a new network or defining long-term aims, it is
more expedient to design a supply system optimized for DG
[6]. The development of such a greenfield network is
demonstrated for the scenario 4. At first the future
development of the area of interest has to be appraised, as
done in the study with the scenario 4. The difficulty in
planning a network with DG is that the layout has to be done
for two contrary extreme conditions. One case is maximum
DG at minimum power consumption and the other one
minimal DG at maximum power consumption. In both cases
an optimal operation has to be guaranteed for all network

CICED2008

Technical Techn ical Session 5

Development Planning of Distribution

conditions. Therefore different basic principles, regarding


voltage quality, loadings, protection and general network
design have to be elaborated for both cases. Depending on the
corporate philosophy these principles can be decided
individually. The principles used in this study are illustrated
in Figure 9.
general netwo rlldesign

vo ltage qua li ty

P/Pn[%)
70

60 j-

-+- -+--+- -t-- - --t--,-+--+- + +-- -+::..."

protecti on

rn-n coreoco

D
rlld ,al distribution system

the voueaeincrease because 01


plants al one node musll'lOl reach

homogeneous IoacI of all

'2 %

transformers

12:00

Figure 10: Transformer loadings (summer day)

D
the voltage increase at maxim um
DG without power consumpnon
muslnotreach+10 %

Figure 9: Basic principles for designing a greenfield network


The basic concept fort the new network is to react to the
DG with a high number of small transformers located across
the area. The greenfield network is composed of eleven small
subnetworks, each a radial distribution system supplied by
one transformer. The nominal apparent output of all
transformers is now at 3630 kVA, a little bit higher than in
the former network. Because of this new structure, consumers
and plants are closer to the transformers so that voltage
fluctuations are much lower and electrical losses are as low as
in a meshed distribution system. Furthermore, cables with a
smaller profile can be used. By uniting several or even all
subnetworks at cut-off points, it is possible to react to
extraordinary situations and (n-l)-operation is so possible.
Concerning to this structure the protection layout was done
for each subnetwork as well as for each combination of the
subnetworks. As a result of this, the layout is much more
difficult. As the mixture of different consumers is very low in
the subnetworks several consumers have a strong effect on
the loadings of the electrical equipment and cables. In Figure
10 this effect is shown by the loadings of the transformers.
All in all the investment costs for the greenfield network
of 0,4 km2 amount to 2 million Euro (low and medium
voltage).
Table 1 outlines the results of scenario 4 and different
network configurations by means of two days.

Conclusion

A considerable increase of co-generators causes rising


loads of the electrical equipment and cables. As well the
compliance of the (n-l)-criterion and the dimensioning of the
fuses are much more sophisticated. Also there is a high
feedback of power and voltage tolerances cannot be
guaranteed for every operating condition. However the supply
of capacitive reactive power, especially by the co-generators,
can help to keep the defmed conditions for voltage quality
and short circuit currents.
Even in a network with many local supply systems, it is
often sufficient to do some selective touching up to guarantee
a supply of high quality. Alternatively it is also possible to
react to the distributed generation with a high number of
small transformers located across the area. The influence of
co-generators commonly exceeds that of solar energy plants
so that the integration of a high number of solar energy plants
into the low voltage network is widely unproblematic in
contrast to the integration of co-generators.
Acknowledgements
This paper is supported by the Institute of Electrical
Power Systems, University of Erlangen, and the division
strategic network planning of the Erlanger Stadtwerke AG
(public utility company ofErlangen).
References

Table 1: Results of scenario 4


nominal operation
abnonnal operation shortcircutt othercriterions
day
type 1p~/P" u m.. Urnin W""
u~
S/I. " max S,,"min protection n-l
Urnin
%
%
%
IINA IINA
% kWh
%
yes
meshed
51 104,0 95,6 720
107,4
95,9 20,6
1,1 no
winter radial
58 103,7 95,1 861
110,9
95,7 14,4
1 no
es
reenfield
81 103,1 95,6 740
108,7
96,4 13,8
0,8 es
es
meshed
42 1031 975 342
1074
978
206
11 no
es
summa radial
48 1037 966 378
1109
978
144
1 no
es
reenfield
65 104,2 96,8 406
108,7
97,8 13,8
0,8 es
es

[1] DIN EN 50160. "Voltage characteristics of electricity


supplied by public distribution networks; German
version EN 50160: 2007" Deutsches Institut fur
Normung e.V., Germany, 2008.
[2] T. Keil, J. Jager, A. Shustov, T. Degner. "Changing
network conditions due to distributed generation systematic review and analysis of their impacts on
protection, control and communication systems", 19th
Cired, Austria, May 2007.

CICED2008

Technical Technical Session 5

Development Planning of Distribution

[3] G. Mehlmann. "Entwicklungsszenarien und netztechnische MaBnahmen zur Erhaltung der VersorgungszuverIassigkeit und -qualitat in elektrischen Energieversorgungsnetzen mit dezentralen Einspeisungen", Diploma
thesis at the University of Erlangen, Germany, Februar
2008.
[4] D.
Nestle.
"Energiemanagement
in
der
Niederspannungsversorgung mittels dezentraler Entscheidung", Dissertation at the University of Kassel, Germany, 2007.
[5] Siemens AG. "PSS Sincal V5.3 Hilfe", PSS Sincal,
Germany, 2008.
[6] VBEW. "Tagungsunterlagen Netzentwicklung", workshop of the VBEW in Munich, Germany, 2007.
[7] VBEW. "Tagungsunterlagen Verknupfungspunkt von
EEG Anlagen", workshop of the VBEW in Nuemberg,
Germany, 2006.
[8] VDEW.
"Eigenerzeugungsanlagen
am
Niederspannungsnetz. Richtlinien filr den Anschluss und
Parallelbetrieb von Eigenerzeugungsanlagen am Niederspannungsnetz", VDEW, Germany, 2001.

You might also like