See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/255989307

Virtual Power Plants-general review: structure, application and optimization

Article · January 2012

CITATIONS READS

149 6,366

2 authors, including:

Jarosław Milewski
Warsaw University of Technology
315 PUBLICATIONS 2,437 CITATIONS

SEE PROFILE

All content following this page was uploaded by Jarosław Milewski on 22 August 2017.

The user has requested enhancement of the downloaded file.

 Open Access Journal

Journal of Power Technologies 92 (3) (2012) 135–149

journal homepage:papers.itc.pw.edu.pl

Virtual Power Plants – general review: structure, application and

optimizationI
Łukasz Nikonowicz∗ , Jarosław Milewski
Warsaw University of Technology, Institute of Heat Engineering)
21/25 Nowowiejska Street, 00-665 Warsaw, Poland

Abstract
The article presents information about Virtual Power Plants (VPP). Numerous papers have drawn attention to
the new concept of generation and management of energy. The VPP concept underpins the growing number
of installed Renewable Energy Sources (RES). The concept of smart controlled Distributed Energy Resources
(DER) merits consideration. This article is a review of some VPP ideas and gives insight into and a general
overview of VPP. Some VPP structure and control methods are described, test fields of VPP are presented
and it ends with a short conclusion about VPP.
Keywords: Virtual Power Plant, VPP

1. Introduction a lack of transmission capacity in the power network

for RES. Further complicating matters is the irregu-
The Virtual Power Plant is quite a new concept.
lar schedule that some RES work to: wind turbines
The VPP idea was born a few years ago and has a
for instance are obviously wind-dependent. This
couple of advantages working in its favor. The main
causes serious problems to the Transmission System
concept is based on a centralized control structure
Operator (TSO). RES might be suitable for installa-
which connects, controls and visualizes a work of
tion on household sites. Capacity of thousands or
distributed generators. Combined heat and power
millions of watts in such RES might eventually rival
generators (CHP), fuel cells (FC), photo voltaics
the capacity which is now installed in wind power
(PV), heat pumps (HP), solar collectors and any
plants etc.
other sources of power and heat might be aggre-
gated and cooperate together in the local area. This VPP provides an opportunity to lower the load in
is a good solution for harnessing Renewable Energy the power network. More power is generated lo-
Sources (RES). At present RES have problems hook- cally and is shared by participants without needing to
ing up to power networks. This happens because of transmit it over long distances at high tension. There-
fore one energy loss factor is either minimized or
I
Paper presented at the 10th International Conference on eliminated. VPP causes a sea-change in energy re-
Research & Development in Power Engineering 2011, Warsaw, lations. The participant is no longer merely a passive
Poland user. Being a part of VPP means everyone involved
∗
Corresponding author
Email addresses: [email protected]
can influence the power system in an active way, al-
(Łukasz Nikonowicz∗ ), though naturally only to a certain extent: it does not
[email protected] (Jarosław Milewski) mean that participants are responsible for switching
 Journal of Power Technologies 92 (3) (2012) 135–149

devices on and off.

Heading the VPP is a computer system controlled
by the Distribution System Operator (DSO). This
could be organized on the basis of an artificial neu-
ral network. In fact, a VPP could be supported by
any household which has as little as 1 kW capac-
ity in a generator such a PV, FC, CHP etc. In VPP
irrespective of how much generation capacity is in-
stalled in a single building, the most important fea-
ture is connecting all the sources together and run-
ning them so as to obtain a state of self-balance in
the most effective way. VPP places more attention on
local generation, meaning that central generation can
operate in more stable conditions. All peaks of heat Figure 1: The CCVPP Design, after [1]
and power demands can be more easily optimized by
DSOs. Storage of heat or electricity should be used
as well. This will help to achieve appropriate condi-
tions of VPP work.

2. Main concept
The term Distributed Energy Resource (DER)
comprises Distributed Generation (DG), Energy
Storage and even Electric Vehicles.
If DER can cope with electricity peaks, it is possi-
Figure 2: The DCVPP Design, after [1]
ble to use power capacity to generate additional en-
ergy off-peak. This energy can then be sold on the
energy market. DERs can be grouped and managed • DCVPP (Distributed Controlled Virtual Power
by a central unit, thereby becoming visible on the en- Plant)—Fig. 2—DCVPP introduces a hierarchi-
ergy market. And it is open to any type of generation cal model by defining VPPs on different levels.
technology. A local VPP supervises and coordinates a lim-
The main focus points within VPP research have ited number of DERs while delegating certain
been: decisions upwards to a higher level VPP. This
design can help simplify the responsibilities and
• feasibility of DER market participation;
communication of the individual VPPs.
• VPP control and coordination optimization;
• FDCVPP (Fully Distributed Controlled Virtual
• design of VPP and power system. Power Plant)—each DER acts as an indepen-
dent and intelligent agent which participates in
2.1. Structures of VPP and reacts to the state of the power system
Three different approaches to VPP can be used: and market. This design holds much promise
• CCVPP (Centralized Controlled Virtual Power as regards supplying a dynamic and optimized
Plant)—Fig. 1—in this design all control logic power system.
lies with the VPP and all knowledge about the
market and the planning of production is sepa- 2.2. Examples of VPP usage
rated from the DER. The advantage of this de- The most significant European projects that in
sign is that the VPP is given a simple way of some way use the concept of VPP and integrated
utilizing the DERs to meet market demand. DER:

— 136 —
 Journal of Power Technologies 92 (3) (2012) 135–149

ized. Only one standard can be allowed and an

information package must be determined.

• interchangeable strategies—the behavior exhib-

ited by a production unit should change depend-
ing on the choice made by the unit’s owner or
VPP operator.

• security and robustness—the system must be

protected from external dangers and must have a
procedure for operating in the event of lost com-
munication.

Proposal solutions:
Figure 3: The FDCVPP Design, after [1]
• loose coupling solution—the ’match maker’
module will be introduced. All information
• SmartGrid; about the DER unit will be entered in the
database of that module. It will be responsi-
• FenixProject—the goal of the project is to move
ble for searching for connections with possible
away from traditional management of small
VPP operators. In the next step, after choosing
units in a power system. Thus, the ’fit and for-
a VPP, the match maker sets up a connection
get’ principle must be rejected. Through the
between a VPP and the unit.
FenixProject all sources will be integrated in an
active way with the system. The new approach • generic adoption solution—all exchange of data
should be used for every kind of DER unit. The should be made according to one standard.
FenixProject tests two types of VPP: commer- XML is a proposal for this task.
cial and technical.
• interchangeable strategies solution—the system
• Ecogrid Project—introduces the concept of the has to react in a quick and efficient way so as
Distributed Energy Market (DEM). The main to dynamically change behavior to fit the situa-
purpose of DEM is to put the end-user at the tion. It must be possible for the VPP operator,
center of the power market and provide the sys- for instance, to change the strategy or logic of a
tem operator with the most cost effective solu- production unit.
tions for system management. An example of
this is functioning on the island of Bornholm. • security solution—the security standards and
specifications for web services must be defined.
In the case of a lost connection between the
2.3. Challenges for VPP implementation DER unit and VPP operator, the DER unit con-
Many different approaches have been taken to im- nects with ’Match Maker’ to gain information
plementation and VPP control. Each has had to cope about new, achievable connection with the op-
with some challenges: erator of another VPP. Then the connection is
established in a dynamic way.
• loose coupling—the production unit owner
should be able to freely choose the VPP with
which he would like to be grouped. The VPP 3. VPP review
group of members is not static.
3.1. European Union 5th Framework Programme
• generic adoption—communication between the In [2] a project is presented that was realized un-
DER units and VPP operator must be standard- der the EU’s 5th Framework Programme. The aim

— 137 —
 Journal of Power Technologies 92 (3) (2012) 135–149

was to develop the VPP concept, to implement and and hydraulic system. The MicroCHP was fed by
test it and show the results. Under consideration natural gas. Heat was consumed on site and the elec-
was whether fuel cells as DER for VPP could be in- tricity first went to the inverter where DC current was
stalled at household locations. 31 stand-alone res- transformed into AC current and then the electricity
idential fuel cell systems were installed. Each unit was supplied to the building’s main network.
had 4.6 kWel and 9 kWth . An Energy Manager was
set up to control the whole system. This module was 3.2. EDISON Project
responsible for benefits for end-users and grid pur- The EDISON Project [3] aims to integrate an elec-
suer. A Central Control System (CCS) was created to tric vehicle fleet with the power system. The inte-
manage all the fuel cell systems. CCS communicated gration is realized by VPP. RES cause problems with
with the on-site Energy Manager and allowed the balancing in the power system. To this end a compre-
utilities to control the micro CHPs in terms of peak hensive solution must be developed. Electric vehi-
demand and defined load profiles. Wireless trans- cles can be treated as energy storage units. The first
mission standards, GSM and radio ripple control re- issue is to draw up a schedule for charging electric
ceivers were used for communication purposes. The vehicles. This must factor in all boundary constraints
project was successful. The whole system was sta- of the system and aim to minimize costs. To solve
ble. There were no emergency cases involving units this problem a special platform must be developed.
being turned off. Fuel efficiencies of up to 90% were It will be coordinated with the power system and en-
achieved was (with 30% electrical efficiencies). ergy market to gain all necessary information. RES
The low temperature PEM fuel cell system worked will be taken into consideration as well. The com-
for 138 000 hours. In that time ca. 400 MWh of elec- prehensive system must supply energy to all electric
tricity were generated. More than 50 million mea- vehicles immediately. All boundary constraints of
surements of data were taken and analyzed. The sys- the electric distribution network must be taken into
tem was tested to check how VPP delivers electricity consideration. The platform described above will
supplies. The results demonstrated that there was no be located on the island of Bornholm. Every elec-
latency time in delivery. There are some problems tric vehicle in every location on the island can be
which have to be solved before developing this type linked with the power system via VPP. A simulation
of system for the mass market: was performed to gauge the influence of a fleet of
electric vehicles on the local electric distribution net-
• costs must be reduced significantly to increase work. One aspect which differentiates the EDISON
the technology’s economic viability; Project VPP from other VPP is the common usage of
• the system must be simplified to improve relia- electric vehicles as active energy storage units. Most
bility; VPPs concentrate only on intelligent management of
generation units. Two possible ways of implemen-
• the temperature of the heat output must be tation are considered. In the first one, VPP will be
increased to become compatible with existing integrated as part of the power system. In the sec-
heating systems, and to give opportunities for ond one, VPP will be a new system that cooperates
tri-generation. with the existing power system. In the second ap-
proach VPP will be a new subject on market. If VPP
The total cost of the project was EUR 8.3 million. is introduced with the power system as is being con-
The fuel cell system comprised a fuel cell battery, sidered, then it will stay part of the power company.
peak heat boiler, hot water tank and control mod- VPP will provide balancing tasks as Balancing Re-
ule. An Energy Manager controlled the whole sys- sponsible Party (BPR). VPP might be a perfect tool
tem. The primary aim of the Energy Manager was to to smooth the boundary between demand and supply.
supply heat energy to meet heat demand in buildings. Standalone VPP architecture is an alternative to the
To do this, it communicated with the CHP system in above. There VPP is BRP too. But it is indepen-
household sites and controlled the fuel cell, boiler dent and works as every other member of the market.

— 138 —
 Journal of Power Technologies 92 (3) (2012) 135–149

It buys and sells energy based on the collected data these tasks require solutions and better models
and state of each generation unit. The most impor- for improved results.
tant task is to create balancing schedules. VPP of
EDISON Project contains 3 main modules: • optimization—various objective functions can
be optimized, i.e. costs, power balancing in
• control module for each DER; the case of intermittent operation of RES units
and power supply for electric vehicles. Opti-
• data collecting module;
mization must be developed to achieve a better
• connection, cooperation and communication global optimum based on local optimums.
module.
3.3. Konwers 2010
Each of the above modules contains other modules.
Basic requirements and experience with regards to
For standalone VPP architecture the whole system is
VPP are set out in paper [4]. Increasing reliance on
more complicated than if VPP is integrated in the ex-
RES causes more problems in terms of balancing.
isting power system structure as part of a power com-
RES depend strong on weather conditions. CHP op-
pany. Bornholm was chosen as test field for VPP.
erations too are driven based on heat demands which
52 DER units are located around the island and 35
are determined largely by weather conditions. Liq-
of them are wind turbines. It is good place to test
uidating the reserve power of central power sources
such a system operating in island mode. 27 000 con-
is justified from an economic point of view. This re-
sumers of electricity are on the island. Total capac-
serve is used to compensate for a lack of power in
ity is 135 MW and maximum load is 55 MW. The
the power network resulting from unpredictable RES
EDISON Project checks how electric vehicles can
generation. It is more sensible to transfer a balanc-
cope with wind farm generation. The potential ex-
ing task to a different level of structure. This struc-
ists to have active management of electric cars with-
ture should contain different types of DERs, energy
out any disruption to car owners. A simulation was
storage units and demand control facilities. It can all
performed based on a model of the power network
be clustered into VPP structure which can perform
of the island. This might be done using commercial
as system power plant. The operations of each unit
software of Matlab/Simulin/Powersin or DigSilent’s
can be scheduled in advance. A distributed energy
PowerFactory, which simulate and analyze transmis-
management system (DEMS) supervises the whole
sion or distribution networks, power sources and
system, taking into consideration all boundary con-
consumers. The model can be used for data manage-
ditions.
ment, prediction and optimization of the operation of
the whole system. With the present model it is pos- All these tasks can be performed because of in-
sible to simulate energy flow in a power network de- novative data transfer methods, communication and
pendent on electric car movement. All calculations remote control which together can monitor a large
are made with 15-minute intervals. The calculations number of distributed energy sources. VPP con-
are used to make an energy flow map. This can be trols all energy flows in the system and factors in the
used to determine where the network and transform- weather forecast.
ers are overloaded. Generated energy and consump- Modern power systems are based on the central-
tion are balanced by regulating wind turbines and ized generation of electricity and/or heat energy.
power plants. Further work will be dedicated to two However, global trends are heading toward increas-
problems: ing numbers of distributed generators. This means
the management process has to adapt to the presence
• prediction—electricity demands must be pre- of distributed generators and their unique method of
dicted in order to create a schedule of power operation. The power system has to cope with un-
source generation. Charging electric cars is an- predictable conditions of operating DG. Therefore,
other issue. Wind conditions must be predicted an innovative approach to management of the power
using weather forecasts and historic data. All system is demanded. The rise in and penetration of

— 139 —
 Journal of Power Technologies 92 (3) (2012) 135–149

RES is a challenge. The new approach to manage- is covered by DERs of VPP. If electricity produc-
ment must be cost effective, economic and provide tion is insufficient, then extra electricity will be de-
a stable operating system. The balancing process in livered to the VPP from external sources. Connec-
some areas can be taken over by VPP. VPP will op- tion to the external power network is necessary. The
erate based on schedules made in offline mode. VPP Kalman filter was used to supply predictions of elec-
will supervise the schedule realization of each DER tricity and heat demand. The algorithm uses histor-
in online mode. VPP can be integrated vertically or ical data—solar exposure, temperature and calendar
horizontally. One VPP system can be a part of an- data as well. Mean absolute deviation varies between
other, bigger VPP system. It is possible to connect 6–14% depending on the type of load. It must be
many VPPs to the existing power system. As can noted that this filter is very sensitive to changes in
be seen, VPP architecture is a very flexible structure. system structure. New units cause increasing error.
This is one of VPP’s biggest advantages. The basic Prediction of RES generation is performed on the ba-
functionality of a VPP is provided by DEMS. The sis of weather forecasts, but its accuracy is very low
DEMS system performs generation, storage and load with a mean absolute deviation of 40%.
management. The main goal of the DEMS system is
to achieve a win-win situation in the power system, 3.4. FENIX Project European project FENIX—
meaning that it will benefit both the power system Northern and Southern Scenarios
and the customers. In Europe VPP is considered as a new ap-
proach to meeting power demand. Two concepts of
3.3.1. VPP—operating description VPP—Technical VPP (TVPP) and Commercial VPP
Renewable generation and electrical and thermal (CVPP) are developed in the project. TVPP is a con-
demand within the supply area is forecasted for cept of aggregated generators which are located in
each 15-minute billing period by offline modules of the same geographic area. DSO is given the real-
DEMS. Based on this, the operating schedule for time local demands of capacity. Those demands can
each DER unit is made for each 15-minute period. be covered by DER. Moreover, the cost and operat-
All schedules are made 1–3 days in advance. Only ing characteristics of each generator are given too.
units with a certain share on the maximum power In other words, TVPP is a local power management
of the VPP are considered. Small units and non- system which gives detailed information about all as-
controllable units are only forecasted. It should be pects of the local system. CVPP has functions which
borne in mind that VPP can be optimized in several contain information about the costs and character-
ways. The control of scheduled operation is made in istics of distributed power sources. CVPP does not
online mode. Unplanned power fluctuations and de- deal with the technical delivery of loads. It is a sys-
viations from the schedule require rapid adjustment tem which enables trading in the energy market and
of the real power flow within the individual period the balancing of trading. TVPP and CVPP do not
by dispatching controllable generation, storage units have to be the same system. TVPP can contain more
and demand in a one-minute time interval. than one CVPP in FENIX. VPP were implemented in
Additional reserve strategies must be provided to two networks. The first of them was the real power
cope with unavoidable prediction errors. It will cover network of Iberdrola in Spain (Southern Scenario)
the reserve power locally with all technical con- and the second was the EDF Energy network in the
straints. The system must stay very simple, compat- UK (Northern Scenario).
ible and complete. An emergency situation should
not eliminate the operation of the whole VPP. In pa- 3.4.1. Northern Scenario
per [4] is described Konwers 2010 project. VPP in- The Northern Scenario concentrates on the usage
cludes CHP fired with biomass, wind turbines, so- of CVPP. It is dedicated to small scale generation: in
lar plant and conventional power plants. All have households and municipal facilities (i.e. civic cen-
to supply households, industry, hotels and offices. ters, conference centers etc.). The main parts of the
The main part of the electricity and heat demand devices are CHP and PV, connected to a low voltage

— 140 —
 Journal of Power Technologies 92 (3) (2012) 135–149

network. But medium scale devices take part in the taining a determined voltage level by providing reac-
test too. The aim of the VPP tests was to prove that tive power to the network; Network Contingencies—
it is possible to use the VPP concept in the present- generators available to work just in case; Tertiary
day network. The idea was to check if loads can be Reserve—power reserves that can be put into the net-
controlled for short periods of time. The VPP system work within 15 minutes and help to cope with imbal-
contains a couple of different structures to fulfill dif- ances; Participation in the Day Ahead Energy Mar-
ferent tasks. The first are generators. Without them ket. As previously, distributed generators can be used
it is impossible to deliver the VPP concept. Next as part of the energy market. Each unit reports their
are control boxes. They provide the data-gathering. state of availability and current level of work and
All generators in the structure tree are visible in the creates a bid. All information is collected and sent
control box system. The system gathers information to CVPP. CVPP processes the received data, makes
about the actual generation level, demands and con- one common bid and submits it to the energy market.
trol flexibility. The data are sent to DEMS hosted After the trading session the returned information is
by DSO and then to CVPP. The solution is prepared delivered to CVPP, which divides it into single bids
in response to incoming information and is then sent and informs the DSOs. The DSOs receive informa-
to generators as returned data. The answer contains tion about the assigned capacity output and a work
information about the best generation level from the schedule for each generator. Then every DSO has
aggregated devices in the context of the actual situ- to make a decision about accepting or rejecting each
ation on the energy market. Consideration has been bid in terms of technical feasibility. The results of
given to involving the DSO as an active participant of this validation from all DSOs are sent back to CVPP
the network in the future development of VPP. The system which hands it over to TSO. The TSO has full
DSO will have more rights to control and to act in all information about the work schedule of each gener-
VPP processes. The main advantage of the system at ator. As a result, aside from the traditional balanc-
hand is the near real-time visibility of generation and ing procedure, a new participant of the transmission
demand in the area where VPP is installed. It should network has been created. This manner of cooper-
not be forgotten that another important value is the ation between DSOs, TSO and other participants of
visibility of flexibility as regards all elements which energy markets is well known in energy systems in
are aggregated in the main layer of VPP—from gen- many countries. The difference is the possibility to
erators to the distributed network. earn money from the hundreds and thousand of units
which operate in the VPP structure. It benefits from
3.4.2. Southern Scenario the rule that the bigger you are, the more you can
In contrast to the Northern Scenario, the Southern do. The Southern Scenario delivers experience in
Scenario focuses on generators which are connected this field. It shows that the energy market is a place
to a medium voltage network. They might serve as where VPP can be used in a commercial way.
an ancillary service to DSOs and TSOs. In the net-
work at hand DER 12 capacity of about 170 MVA 3.5. ’Smart’ Heat Pumps
is installed, which determines about 35% of all ca- The increasing capacity of wind farms causes
pacity linked to this medium voltage network. The problems with balancing generation and demand.
VPP works as a parallel control system and is op- There is self-evidently little ability to affect gener-
erated by DSO. There is no interference in real net- ation. On the other hand, demand can be used as
work movement. All information about the network a factor to cope with problems of surplus capacity
is downloaded in real-time from a SCADA system. in the network. Demand might be shaped in an ap-
In this system no difference exists between DEMS propriate way to meet available generation. This is
and CVPP. DEMS is the same as CVPP. Data are ex- the main idea behind using heat pumps in the en-
changed between CVPP module and control boxes ergy system. ’Smart’ heat pumps can help to balance
by the GPRS network. VPP was used to show its generation and demand and to manage network con-
usefulness in: Voltage Control—support for main- gestion. The role of heat pumps in the distribution

— 141 —
 Journal of Power Technologies 92 (3) (2012) 135–149

network might be enhanced through active partici- units, each unit having 1 kW electric. During tests
pation in the energy market. The smart heat pumps not all of the units were connected to the same low-
idea is being considered in a few countries, i.e. Ger- voltage network, but they all had a common sub-
many, Switzerland and Denmark. A description of station. ECN developed the PowerMatcher module,
the smart program is set out below. which is responsible for coordinating supply and de-
mand in the electricity network. A special software
Germany. Vattenfall Europe launched VPP in and communication module was installed in each mi-
Berlin. VPP controls the operation of CHPs and heat croCHP unit. The software controlled a work of unit
pumps which are aggregated in VPP. In Germany the on the basis of a number of important economic fac-
whole concept is based on energy prices in the en- tors. In such a case there was no need to implement a
ergy market. Currently, the whole system contains central optimization algorithm. Bids took the form of
only 30 heat pumps and CHP in total—20 HP (with information exchanged between units and the Power-
heat capacity of <25 kWth each) and 10 CHP. The Matcher operator—bids as a signal for the electronic
capacity is 30 MW and the goal is to reach 500 MW market. The bids consisted of information about the
by the end of 2011. The system is a response to in- price which is acceptable for the unit’s owner in the
creasing unpredictability due to more and more wind sense of buying or selling electricity (and capacity).
generation capacity. This causes sudden energy price In reply from the market, price signals were deliv-
fluctuations on the energy market. This can be re- ered. An autonomous decision about operating or
duced by using heat pumps. When the energy price not operating and waiting for the next round of bid-
is low, the heat pumps are turned on. When the price ding was taken on the basis of those signals. A Pow-
is high then CHPs are turned on and generate surplus erMatcher module can be connected to DER or other
power. The whole process is controlled by a Vatten- PowerMatcher modules. This is possible due to stan-
fall control center. The VPP structure in Germany dardization of the transmitted information package.
is simple. There are units which generate power or Various units can be clustered onto different levels
heat. Their operating schedule is mainly dependent and there is no problem with connecting units to
on the profile of heat and power demands of cus- PowerMatcher modules. In other words, one Pow-
tomers. These units have a communication module erMatcher can coordinate several units and can be
which sends information about demand. All units connected to another PowerMatcher module—on a
are aggregated by VPP. VPP gathers all information higher level—which in turn coordinates the opera-
flows from units and processes them. Everything is tion of other units etc. The unit’s owner could de-
passed to the control center to ensure quality ser- termine the main goal of operating the unit. Bids
vice and control. After correlation with energy prices are then sent to the PowerMatcher module. The bids
taken from the day-ahead market, the control center contain the aim preferred by the owner.
generates a dispatch schedule for each HP and CHP Field tests were made in 3 cases:
for the following day. The report is then sent to the
units. There is two-way communication between the 1. the load was assumed to be the standard pro-
units and control center. The aim is to achieve an op- file of energy usage in households in the
timized system which meets all demand and is able Netherlands—no micro CHP;
to generate extra income while saving energy. 2. the same energy consumption as in 1, mi-
croCHP units were installed and operated only
The Netherlands. The usage of household mi- in a heat-demand driven manner;
croCHPs was described in [5]. Those microCHPs
3. as in 2, but VPP was launched and microCHP
can be used as sources of electricity and heat for lo-
functioned in a new environment of collabora-
cal demand. 10 units were clustered in VPP and the
tion between VPP and the market.
performance test was conducted. ECN and Gasunie
coordinated this test field. It was found that the op- The field tests were conducted in May 2007. There
portunity exists to reduce substation peak load by was only heat demand for hot water. All house-
30–50%. Stirling engines were used as microCHP holds had a 120 liter tank for hot water. The paper

— 142 —
 Journal of Power Technologies 92 (3) (2012) 135–149

but implementation with other, different units does

not pose a problem and should not cause any errors
in the VPP system.

4. Methods of VPP control

4.1. Ohmic resistance control method

The Virtual Power Plant can be considered as a
cluster of distributed electricity generators (DEG).
They are attached to the telecommunication net-
work. In the paper [6] the possibility of VPP operat-
ing without additional costs of transmission are de-
scribed. MicroCHPs which generate electricity and
Figure 4: Field test result of clustered control of 10 microCHPs heat are considered. They have a high total effi-
after [5]
ciency. The typical heating system for a household
comprises a CHP integrated in the central heating
showed results for one day during which only 5 mi- system and an additional burner which is used dur-
croCHP units were participating without any prob- ing winter peak time, plus a hot water tank. The
lems and disruption. Figure 4 shows total electric- electricity generated can be used on site or trans-
ity demand during one day. As can be seen, there ferred by a low voltage network to the power system.
were 4 peaks. PowerMatcher reacted to those peaks, When there is insufficient generation of electricity
shifting additional power production. The field test and unbalancing occurs, then external electricity is
showed that 5 microCHPs mitigated the maximum supported by a low voltage network. Two modes of
value of peaks. But the boundary factor that re- operating are defined - heat driven and power driven.
sulted in a poorer quality outcome was the lack of Hot water storage gives extra capacity for storing
sufficient heat storage space: units could not produce heat, which can be used to improve the operations of
more electricity without producing additional waste the whole system, and for switching between oper-
heat. The hot water tanks were designed to cope with ating modes in light of the heat storage status. Each
the waste heat problem, but they had limited capac- kind of system has specific parameters. The ohmic
ity. And in a situation where peak loads appeared resistance of a low voltage network is higher than
in very short intervals there was insufficient free ca- that present in a mid or high voltage network. This
pacity to take over additional heat production. Hot fact can be used to define areas where the power load
water tanks were filled during the first peak. The pe- is the greatest. Such areas are characterized by high
riod of time between the first and second peak was voltage drops. Thus, the signal can determine the
very short, so there was no time to use the heat stored load of each CHP unit in the VPP system. This re-
in tanks. As a consequence, the engines could not lates to units which operate in power driven mode.
operate. All tests collected information about VPP The idea is to increase the generation of CHP units
potential. They clearly demonstrate that the DER in- in cases where the voltage in the feed-in nodes drops
telligent management system produces benefits. The below certain levels and vice versa.
current way of power system management—fit and
forget—delivers more system load. The suitable op- 4.2. Marginal costs
erating of DER units can result in lowering and mit- Supplying power demand to balance the power
igating the demand curve in the system. system in real time is one of the key advantages of
The authors claim there is no problem with group- VPP. The reserve power is at TSO’s and DSO’s dis-
ing together different types of power sources. The posal. VPP manages all units in the most cost ef-
case presented above featured Stirling engines only, fective way and minimizes total balancing costs. To

— 143 —
 Journal of Power Technologies 92 (3) (2012) 135–149

operate correctly some input factor must be deliv-

ered to VPP. The impulse can take the form of the
marginal electricity cost of the individual DER units.
Marginal electricity costs are highly dependent on
the local context and change over time. For exam-
ple, CHP generates electricity depending on heat de-
mands. The more heat is needed, the more electricity
is generated and vice versa. VPP is a structure which
contains a number of different types of DER. For the
typical situation the power of each unit is not very
Figure 5: Structure of the prognosis based optimization algo-
high and so the DER units are small. The production rithm for cogeneration system, after [8]
of energy is a dynamic process, so the marginal en-
ergy costs are dynamic too. In article [6], a method
is presented to determine DER marginal costs and are controlled by price signals—if the energy
benefits achieved through using bid strategies. DER price is low then the storage device runs and
units have to send information about marginal costs vice versa.
to VPP. This information is sent in the form of bid- • mixed strategy—this can be used for CHP units
ding formulas or demand curves which determine with additional heat sources and heat storage
the needs of VPP for electricity at established prices. units. It is hybrid of the above two methods.
Negative values in the curve diagram means the DER In paper [6] a bidding strategy is presented for
unit is able to produce extra power at set price level. different types of DER.
Bid offers are made by an agent using special soft-
ware which is able to create a complex bid schedule 4.3. Optimization of CHP
for a particular moment. These offers are made based The algorithm for CHP optimization is described
on: in [8]. The aim of the algorithm is to set all devices
in such a way as to maximize the benefits from sys-
• current operating state of the DER unit;
tem operation. Everything is based on the variation
• economic parameters such as marginal operat- of electricity prices over time. The operation sched-
ing costs; ule of units must be known one day before physical
supplies. This schedule has to be delivered to EEX.
• market environment with all market mecha- Underpinning the algorithm are the predicted elec-
nisms; tricity prices on the spot market and the heat demand
forecast. Heat can be produced by boilers, CHP or
The author of [7] sets out 3 different strategies which a heat storage tank. The heat storage tank gives an
might be used in the bidding process: opportunity to separate the production of electric-
• fully marginal-cost based strategy—this can ity from heat. A CHP can operate even if there is
be used in a situation where the VPP sys- no demand for heat from the user. The main need
tem contains only distributed electricity gener- is to maintain heat supply to a customer at an ade-
ators. It considers fuel prices, unit efficiency, quate level. The CHP can operate in such a way as
running-history dependent maintenance costs to generate surplus electricity which can be sold on
and startup costs. the market. The algorithm maximizes benefits from
sales of surplus electricity and handles heat supplies.
• fully price history based strategy—this can be At the same time there are boundary conditions on
used when storage devices are used, e.g. ac- the algorithm objective function. The boundary con-
cumulators. It considers minimum and maxi- ditions come from the technical restriction of oper-
mum prices from previous time periods and the ating CHP units. To achieve the best fitted result of
level of available storage space. Storage devices optimization, prediction data must be delivered. The

— 144 —
 Journal of Power Technologies 92 (3) (2012) 135–149

authors suggest using statistical methods with em- lems are located exactly on the low or mid voltage
pirical functions to fit loads (prognosis with multiple network level. They manifest themselves in: change
linear regression). Physical model based functions in energy flow direction, overload of network, prob-
can be used. In the opinion of the authors, neural lems with frequency and balancing. The control sys-
networks are not well-suited to make forecasts. In tem of VPP and optimal structure of VPP are de-
the method employed the forecast is done using a scribed in paper [9]. The authors highlight need for a
database with a time horizon of at least 1 year. The central Energy Management System (EMS) module.
data used are: weather conditions, calendar data and Each VPP has to be connected directly or indirectly
the commercial weather forecast. As a result the heat with EMS so as to enable the exchange of important
demand curve is obtained for a specific building. The data. The EMS controls the entire exchange proce-
results correlate reasonably well with real data and dure in real time. The data contain information about
values. The idea of optimization is to find the max- the current situation and state of each market partic-
imum or minimum value of some objective function ipant. The appropriate communication must be pro-
in some calculation area. The authors used the MILP vided to transfer data. The VPP system should be
method to optimize the objective function (Mixed In- local in nature, react quickly with minimal latency
teger Linear Programming). This method can find time and have the ability to connect new units. The
the optimum in a model which is described by linear network hierarchy structure should adjust according
function as well as integer variables. Optimization to the number of users. Too many points connecting
calculations for a local heat system are formulated with the EMS at any one time may result in the whole
in the solver-independent language AMPL and are system slowing down and thus overloading. The au-
solved by CPLEX optimization software. A couple thors point out problems with energy flow. In the
of formulas which are the mathematical models of considered VPP more attention must be put on con-
different elements of the energy system are presented trol of the energy flow in the network. Thus, the mea-
in the paper. There are formulas for the heat system, surement of energy flow has to change. There is at
CHP, boiler and heat storage. The objective function present only measurement on the power sources side.
is determined as maximum benefit from total heat The new concept assumes measurement will take
and electricity sales considering the costs of energy place in real time on the demand side too. Measure-
production in each CHP and boiler unit. The max- ment devices will have to be installed and a special
imization of benefits can be contrary to other goals, protocol for data transmission implemented too. The
e.g. to minimization of primary fuel energy. The authors suggest using a measurement system coupled
model based on MILP was used to optimize the op- with GPS technology. Transmission will be possi-
eration of an existing heat system which is supplied ble by satellites. The operator demands full possible
by [3] the same CHP units. The forecast was done knowledge about the network to determine what hap-
using data from the EXX spot market of 2004. The pens in individual lines of the network. Several ser-
simulation was performed and compared with real vice tasks must be performed to avoid network faults;
values. 2 power diagrams were obtained—one with usually the TSO will be responsible for this. Through
the simulation and other with real operation. The to- this activity the system can operate without distur-
tal analysis for the whole year showed that there was bance. The role of the DSO is very limited. This sit-
a cost savings potential of 10%. Moreover, the ap- uation will change due to VPP presence. The DSO
propriate usage of heat storage reduced the energy will cooperate with VPP and more service tasks will
losses to a small extent. As is shown, great potential be transferred to the lower level—that of the DSO.
for optimal management still exists. The power system operation schedule will be done
by DSOs. They will deliver information about the
4.4. Reduction of generation costs reserve power of the DER for the following day.
The emergence of new sorts of power sources such
as RES causes new problems. Due to their connec-
tion to the low or mid voltage network, those prob-

— 145 —
 Journal of Power Technologies 92 (3) (2012) 135–149

Figure 6: Structure of the VPP, after [9]

Figure 7: Flow chart of the optimization control, after [9]

4.4.1. Optimization approach 5. Decision-support software for VPP manage-

ment
The objective function of EMS is determined as
minimum: energy generation costs, total costs and The decision-support tool for Virtual Power Pro-
transmission losses. The boundary for the objective ducers is presented in article [10]. A tool usage anal-
function is network overload. Measurement technol- ysis is made. The authors highlight that long dis-
ogy must be improved, as mentioned above. The tance energy transmission is unjustified in terms of
considered case involved a VPP system with 3 en- DER production. All consumption of the produced
ergy suppliers and 4 different consumers. A desalina- energy takes place at or near to the site of produc-
tion plant and electrolyzer for hydrogen production tion. The DER units are connected with the distribu-
were implemented as consumers. It clearly demon- tion network. Due to the increasing number of RES
strates that VPP also has great potential for operating there will be a marked increase in energy production
with different kind of units. Nowhere else in the liter- from such decentralized sources in the near future.
ature was a solution of this type described. EMS lies Thus, a solution needs to be designed that will be
at the head of the whole system and directs all energy able to make effective use of that renewable energy.
flows to fulfill the needs of the consumers and to op- To achieve that aim, broad-based knowledge must be
timize use of the generated energy. The entire model gathered on all kinds of DER units. The main idea
was tested using data from the German energy sys- is to develop a system which provides an opportu-
tem. A simulation of the VPP system was performed. nity to coordinate all units in one common system
It transpired that there is a specific capacity of wind without sharing DER units. All units must be able to
farm for which total operation costs of VPP are min- cooperate together to achieve the common aim. The
imal. For the data used, this capacity was 125 MW required knowledge concerns the particular operat-
level. More information about structure and test re- ing features of units such as: technology ripeness,
sults are described in [9]. profitability, availability, reliability, production and

— 146 —
 Journal of Power Technologies 92 (3) (2012) 135–149

capture of greenhouse gases, relation with external were conducted. The first one showed that VPP was
factors and lifetime. able to cope with the problem of a sudden turning
The schematic diagram of VPP is presented in the off of the wind farm. The second test simulated the
paper. It consists of 5 layers which are closely inter- situation where the balancing from the VPP side was
connected. They are: meteorology, power producer, not at an appropriate level, so some unbalancing oc-
VPP, Market and System operator. Fig. 8 shows the curred in the distribution network.
challenges facing VPP and the operations required Matlab/Simuling was used to make ViProd.
to achieve results. The schematic gives a general The following conclusions were presented by the
overview on VPP foundations. authors. The appropriate designed VPP structure can
The name of the tool is ViProd. It is responsible eliminate:
for VPP simulation and its interaction with market. • uncertainty of energy production forecasts;
This tool uses the characteristics listed above and
helps with the decision-making process. It has been • fines for unbalancing;
split into 2 parts. The first part calculates the en- • the lack of small energy producers on the
ergy production for the one day ahead and the second market—there is space for small players;
one simulates generations. It is possible to simulate
energy production from wind turbine, small hydro- • problems with the CO2 market;
electric and photovoltaic plant. Real external data is • high management costs.
needed to make correct calculations, e.g. wind direc-
tion and speed, water flow rate, temperature or solar The following step will be to connect ViProd to the
radiation coefficient. energy market simulator—MASCEM.
The tool simulates the energy production possibil-
ities for each type of DER unit, taking into consider- 6. Conclusions
ation the characteristics of local demand, peaks and The number of small DER units will increase over
time variation of factors which can influence produc- time. Currently, the world is powered by large power
tion. Prediction about factors gives approximate real plants which supply power systems. But this situa-
conditions and reduces forecasting error. The arti- tion will change due to RES among other things. The
cle presents only the part of the tool which forecasts central power sources will be turned into decentral-
wind condition based on various input data and the ized sources. This trend is observable in Denmark,
part of the tool which simulates the operation of wind where the number of small CHP units has increased
turbines. The simulator uses results which are de- significantly in recent years. To adapt to change, a
rived by the first part of the tool, and information suitable management system will be needed. VPP
about possible energy generation over a defined pe- is a concept which is able to cope with managing a
riod of time is delivered as a result. The application large number of different types of DER units. The
is responsible for running all clustered units in one biggest advantage of VPP is its modular structure. It
specific VPP structure. Power is distributed properly can be connected to power systems and comprise a
among a number of DER units in function of sold en- number of DER units. Depending on requirements,
ergy, production cost and available power. A check- extra modules can be added, in order to optimize
ing procedure is performed regarding the predicted the system, secure transmission and/or report results.
amount of energy and power. Next a report is created VPP has the flexibility of building blocks.
about the reserve power of each unit which is at the As is reiterated in the paper, every VPP system dif-
operator’s disposal. A software test was conducted. fers in the detail. There are many approaches to the
Ten units were implemented as the VPP system. The VPP concept. Every scientist and engineer involved
test was performed in order to determine whether the has her/his own vision of VPP. But the core idea stays
reserve power is sufficient to cope with unbalancing the same—gather together and manage as many DER
in the power network. Unbalancing might be caused units as possible to achieve better, more cost effective
by, for instance, turning the wind farm off. Two tests and environment friendly results.

— 147 —
 Journal of Power Technologies 92 (3) (2012) 135–149

Figure 8: Schematic representation of the functioning of VPP, after [10]

— 148 —
 Journal of Power Technologies 92 (3) (2012) 135–149

References
[1] M. Braun, Virtual power plants in real applications.
[2] K. Kok, Short-term economics of virtual power plants, in:
20th International Conference on Elecricity Distribution,
2009.
[3] P. B. Andersen, B. Poulsen, M. Decker, C. Traeholt,
J. Oestergaard, Evaluation of a generic virtual power
plant framework using service oriented architecture, in:
2nd IEEE International Conference on Power and Energy,
2008.
[4] G. Kaestle, Virtual power plants as real chp-clusters: a
new approach to coordinate the feeding in the low voltage
grid., in: 2nd International Conference on Integration of
Renewable and Distributed Energy Resources, 2006.
[5] A. Dauensteiner, European virtual fuel cell power plant -
management summary report.
[6] H. Morais, M. Cardoso, L. Castanheira, Z. Vale, I. Praca,
A decision-support simulation tool for virtual power pro-
ducers, in: International Conference on Future Power
Systems, 2005.
[7] B. Roossien, M. Hommelberg, C. Warmer, K. Kok, J.-W.
Turkstra, Virtual power plant field experiment using 10
micro-chp units at consumer premises, in: CIRED Semi-
nar 2008: SmartGrids for Distribution, 2008.
[8] C. Binding, D. Gantenbein, B. Jansen, O. Sunstroem, P. B.
Andersen, F. Marra, B. Poulsen, C. Traeholt, Electric ve-
hicle fleet integration in the danish edison project - a vir-
tual power plant on the island of bornholm (2010).
[9] L. Sugden, ’smart’ heat pumps: Enablers of low carbon
future, A Delta Whitepaper (2011).
[10] P. Lombardi, M. Powalko, K. Rudion, Optimal operation
of a virtual power plant, in: Power and Energy Society
General Meeting, 2009.

— 149 —

View publication stats