Académique Documents
Professionnel Documents
Culture Documents
2, MAY 2012
Abstract—This paper describes the current connection regime DNOs in terms of power flow directions and magnitudes. This
for distributed generation (DG) in the U.K. and presents a novel ap- uncertainty is currently addressed by relatively conservative
plication of the optimal power flow (OPF) technique for automatic planning rules that mitigate these risks and ensure security of
power flow management (PFM) to manage thermal constraints in
distribution networks. OPF formulations have been used, in an of- supply. DNOs are looking towards active network manage-
fline mode, as a power system planning tool for several years. The ment (ANM) schemes [1] based on monitoring and control of
novel implementation of OPF for “corrective” PFM in an online the distribution network and connected users to increase DG
operational mode, for MV distribution networks, is presented and connection while minimizing, deferring, or even avoiding any
tested in this paper. The authors demonstrate, through simulations capacity upgrades associated with DG connection. ANM can
conducted on a commercially available substation computer, that
such an application of OPF can represent first on, last off gener- thus support renewable generation connections and, therefore,
ator connection agreements that reflect the current principles of help to meet renewable targets through incentives such as the
access in the U.K. Two case study networks, a 33 kV and an 11 kV, renewable obligation certificate (ROC) scheme in the U.K. [2].
provide the basis for assessment of the OPF-based PFM algorithm A prominent approach to delivering ANM schemes is the de-
in terms of computation time to arrive at a solution in the event of velopmental Autonomous Regional Active Network Manage-
a network thermal excursion and the level of DG curtailment nec-
essary to meet network thermal limits. Assessments are made and ment System (AuRA-NMS) developed in a major research, de-
fully discussed of the suitability for an OPF-based approach for velopment, and demonstration industry oriented activity. The
distribution network management within an online network con- core functionality of AuRA-NMS, categorized by industry re-
trol scheme including discussion of the important consideration of quirements, is to perform automatic restoration, voltage con-
control robustness. trol and power flow management (PFM). This differs from ex-
Index Terms—Active network management, distributed control, isting ANM solutions due to the package of functionalities of-
distributed generation, optimal power flow. fered. Thus, AuRA-NMS is a move away from bespoke ANM
solutions to a generic one with multiple functions being hosted
on an open, flexible, extensible platform [3], [4]. Flexibility is
I. INTRODUCTION
required to allow for deployment on topologically diverse net-
works and by being easily re-configurable within any network
different generating technologies has been previously reported The AuRA-NMS solution addresses the complexity issues
and applied in [8] and [9]. of constraint management schemes by hosting generic and dis-
In this paper, the impact of DG on distribution network power tributed functionality that is capable of offering responsive mod-
flows as well as a suitable OPF formulation to address DG ac- ular control actions for any thermal violation, voltage excursion
cess rights is discussed initially. Then, the practical applica- and post-fault restoration. Taking an active approach to man-
bility and the ability of the OPF to operate within a real-time aging power flows ensures that the thermal limitations placed
and closed loop control environment to deliver PFM is demon- upon the network are met while maximizing DG access under
strated. As discussed in the following sections, this practical im- all operating conditions.
plementation was carried out by connecting outputs of the Pow-
erWorld tool with a dedicated PC acting as the real-time sim- B. DG Connection Principles
ulator and ABB’s COM6xx series substation automation com- The connection principle, for “non-firm” generation, prac-
puting hardware. The results of tests on two different networks ticed in the U.K. today is a last-in, first-off (LIFO) disconnec-
are presented and conclusions on the applicability of OPF in this tion order. This requires the most recent connected DG unit to
application are drawn. reduce its output first during a network constraint situation to re-
solve the constraint and return the network to a normal operating
state. Where this action is not sufficient to resolve the problem
II. IMPACT OF DG ON DISTRIBUTION NETWORK POWER the next most recent connected DG units would be called upon
FLOWS AND THE REQUIREMENT FOR ANM SYSTEMS to resolve the constraint, in succession, until the disconnection
list is exhausted. Conventional system planning rules would en-
A. Constraint and Power Flow Management of DG sure that, with no DG operational, the network limits would not
Connections be violated under normal operating conditions.
With an increase in the number and capacity of connected
Network power flows are influenced by the characteristics DG units, the LIFO access approach may hinder some invest-
of power injection and consumption, as well as the impedance ments in DG as newly connected DG units may face poor access
of the connecting circuits. By introducing DG units into dis- through more frequent disconnections. It is important to note
tribution networks, there is a move away from the traditional that LIFO arrangements had an important and positive role in
unidirectional power flow that distribution networks were origi- the initial stage of DG investments (clarity over network access
nally designed for. The resultant bidirectional power flows bring rights and protection of prior network connection contracts) but
added uncertainty to the direction and magnitude of flows in a with a growing need to increase the proportion of electricity gen-
distribution network and may pose significant risk to network erated by renewable low carbon technologies, in order to fulfil
security. obligations from the EU 20-20-20 target [12], the authors ac-
Traditionally network connections for DG have been assessed knowledge that DG connection principles must evolve. Other
against the available network thermal capacity as well as voltage connection principles are being investigated and showing im-
and fault current constraints. For example, an interconnected proved network access and subsequent energy yields [13].
network that has the capability to export excess DG may have The objective of this paper is to explore the role that an OPF
connection agreements that are approved as either “firm” or solution can play in active management of DG within the cur-
“non-firm”. These DG units with “firm” and “non-firm” rights rent commercial framework. The purpose of this is to better
can deliver power to load and export power to the adjacent net- utilize network assets and increase renewable DG output while
works uninhibited under normal network conditions. A “firm” recognizing the future potential and extensibility of an OPF so-
connection would be calculated on the n-1 maximum export ca- lution to deliver alternative principles of network access (e.g.,
pacity (for the highest rated interconnecting line being out of minimize DG set point change or delivering DG access bidding
service) plus the network minimum demand. This generation or sharing). This exploration of the potential of OPF to deliver
can therefore remain connected during an n-1 event and at any ANM is in line with the European Smart Grid Technology Plat-
load demand level hence the “firm” connection designation. form’s [14] interest in building up a “toolbox of proven tech-
“Non-firm” connections are calculated using the maximum niques that can be quickly and economically deployed”.
export capability (the sum of all interconnecting circuits) plus The analysis presented here is based on an OPF formulation
minimum load minus the connected “firm” generation. In the in which the cost of each DG unit reflects its connection order
event that one of the interconnecting lines is out of service with the highest cost of generation associated to the last DG
(during fault or maintenance), the “non-firm” generation would unit connected. Hence, similarly to the LIFO arrangement, this
need to be removed from the network, via a signal sent by the OPF-LIFO approach will first seek to constrain the last con-
DNO. The use of minimum load ensures that connections are nected DG units. In the case where a network is heavily con-
based upon a worst case scenario [10]. gested this OPF based method will have sufficient flexibility
The above discussion indicates that there is a need to develop to change the order of disconnection and, regardless of their
new ANM schemes to enable enhanced DG network access in an LIFO order, leave online the DG units whose contribution to
economically attractive manner. The assessment of one ANM the congestion is very weak and thus the curtailment of their
scheme [11] shows that the lost revenue from carefully con- outputs will not help in relieving congestion. The formulation
trolled curtailment is less costly than the expense of network re- of this described OPF-LIFO approach is derived by modifying
inforcement. The extent of this benefit depends on the network cost curves in the traditional OPF problem, as outlined in the
characteristics and the level of DG connection being considered. next section.
792 IEEE TRANSACTIONS ON POWER SYSTEMS, VOL. 27, NO. 2, MAY 2012
III. APPLICATION OF OPTIMAL POWER FLOW within the OPF formulation however, due to the distributed con-
FOR POWER FLOW MANAGEMENT trol architecture of this ANM scheme voltages would be man-
aged by separate voltage control functions and coordinated ap-
A. Standard OPF Formulation propriately within the control architecture.
Optimal power flow [15]–[17] is a well established technique
used for power system operation and planning problems that B. OPF-LIFO Formulation for Power Flow Management of
can be formulated as an optimization problem. Typically, the Distribution Networks
goal is to minimize system operation costs or maximize social The OPF formulation defined by (1)–(6) is typically used for
welfare, subject to system operating constraints such as power operation of high voltage transmission systems with larger gen-
balance, generation loading limits and network capacity con- erating units. This is a complex problem with a large number of
straints. There are, however, various modifications of the ob- decision variables. The system operator has a set of well estab-
jective function formulation such as minimization of original lished rules and ancillary service arrangements to help maintain
control schedule deviation or bilateral contact curtailments [18]. secure system operation. These rules are reflected in the above
In addition, OPF has been investigated for evaluating the max- OPF formulation.
imum connection capacity of renewable DG units [19] and as As discussed in Section II, power flows at distribution levels
a real-time strategy for minimizing load curtailment to remove are evolving from being one-directional to bi-directional. Since
thermal congestion [20]. However, those previous applications the distribution networks were not designed for such an opera-
of OPF do not consider its use in a real-time generator con- tion, this can lead towards a more uncertain and congested sit-
trol situation, as is the focus of the application described in this uation than previously. This, therefore, requires novel network
paper. management solutions to facilitate increased DG connection. As
Typically, the OPF problem seeks to minimize system oper- such, DG developments at distribution levels are relatively re-
ation costs. In vertically integrated system operation, costs re- cent and thus the role of the DG in distribution network opera-
flect true costs of operating a unit, while in systems where elec- tions is less defined than it is for the generating units in the trans-
tricity markets are introduced, minimization is based on gener- mission system. As a result, effect of DGs on network calls for
ating unit offers submitted to a centralized market. operation techniques that will better reflect these uncertainties
The OPF formulation used within this paper is as follows: and bi-directional flows.
The aim of the OPF-LIFO approach proposed in this paper
(1) is to offer a solution that will support better operation of such
distribution networks. Since LIFO arrangements are currently
subject to applied by the U.K. DNO operators, the proposed method seeks
• Power balance equations to closely follow this LIFO access principle for the majority of
operating conditions. However, it will deviate from LIFO only
to avoid disconnecting or curtailing an unnecessary level of gen-
(2)
erator output. In that way application of OPF-LIFO will allow
for better utilization of DG resources and network infrastruc-
ture.
(3) The proposed OPF-LIFO method is based on the above OPF
formulation (1)–(6), with the only change being in the defini-
tion and formulation of the individual generation cost terms,
• Generation limits
, in the objective function (1). While in the general
(4) OPF formulation these terms reflect either true costs associated
with the electricity production, or the offer curves that genera-
(5)
tors submit when selling energy on the market, in the OPF-LIFO
• Thermal limits formulation terms reflect the order of the generators’
connections. Thus, each generator is assigned a constant cost
(6) function defined as
In the objective function (1), is the cost or offer (7)
function of a generator at bus , and are active and re-
active generation outputs while and are active and re- where is a constant allocated to each generating unit to re-
active demand at bus , , , and are the flect its connection order. This means that in order to priori-
upper and lower active and reactive generating unit limits. tize the utilization of generators that were connected first, they
and are the active and reactive power flows in the line be- will be assigned the lower value of , while those connected
tween and and is the apparent power thermal limit for last will be assigned the highest values. Note that the constant
the circuit (including transformers) between buses and . Note in (7) does not reflect actual generation costs (or offers),
that the power flow limitations (6) are verified for both the re- but, as mentioned above, is rather used to indicate a connection
ceiving and the sending end of lines. Traditionally the upper order which would be considered for resolving any network con-
and lower voltage magnitude limits at node can be included straint.
DOLAN et al.: DISTRIBUTION POWER FLOW MANAGEMENT UTILIZING AN ONLINE OPTIMAL POWER FLOW TECHNIQUE 793
V. TEST ENVIRONMENT
A closed loop test environment was established to evaluate
the performance of incorporating OPF into an online PFM al-
gorithm and comprises the following elements: Substation com-
puter, real-time simulator (dedicated PC), PFM and OPF soft-
ware and the case study network models. Each element of the
test environment is introduced in the following sections.
Fig. 2. 11-kV radial system.
A. Hardware Test Platform
The substation computing hardware used for the development
and testing was ABB’s COM6xx series substation automation
product. The COM6xx is a Windows XP Embedded industrial
computer that offers a powerful substation computing platform
onto which ANM and other control functionality can be de-
ployed.
The COM6xx acts as a gateway for mapping, monitoring,
control and protection signals from remote devices (generators,
transformers, protection relays) into the substation as long as the
protocol is supported by the COM6xx. A full description of the
ABB COM6xx series hardware can be found at [23].
B. Real-Time Simulator
A means of emulating the data transfer from network intelli-
gent electronic devices (IEDs) to the PFM algorithms was de-
veloped. The real-time simulator software runs on a dedicated
PC and calls upon data files that hold load and generation pro-
files. These profiles are used to represent the system conditions
Fig. 3. 33-kV interconnected system.
over time. The network model embedded within the COM6xx
is set to update once each second.
Within the real-time simulator are embedded controller
D. Case Study Network models
functions that simulate network control responses and include:
tap-changers; circuit breakers; real power set-point control Network case studies were used to represent two different real
responses; and power factor set-point control responses. For distribution systems in the U.K. The first case study network is a
the purpose of this investigation only real power set-point section of 11-kV radial distribution network with two DG units
control (that includes DG unit power output ramp rates) for DG (Fig. 2). Common planning practice for 11-kV networks entails
trim/trip controls were utilized. tapered capacity feeders that meet the feeder loading, voltage
drop and post-fault reconfiguration requirements with strategi-
C. PFM and OPF Software cally placed open points to achieve necessary levels of network
security. This tapering limits the prospective DG connection ca-
To achieve network management with the OPF-LIFO pacity on the 11-kV circuits.
approach, a new algorithm was developed. The algorithm iden- The second case study network is the 33-kV interconnected
tifies thermal constraint violations, updates DG cost models and network depicted in Fig. 3. This network is supplied via two
calls upon the OPF engine to solve the congestion constraint 132-kV circuits from “Sub X”. These terminate into two
while following connection and operating constraints. The 132/33-kV transformers at “Sub A” and “Sub B”. These two
outputs of running the OPF-LIFO algorithm are control signals substations operate as an interconnected group at 33 kV via
to the corresponding generating units with updated set points. three circuits. A further two 33-kV circuits radiate to supply
The PowerWorld power systems analysis software package loads and generation and these circuits have normally open
provided the functionality to solve the OPF problem defined points midway along their lengths at “Sub E” and “Sub G”.
above. Additional developed software provides the algorithms Combining the COM6xx platform with the real-time sim-
for congestion detection, cost models representing LIFO and ulator, the OPF-LIFO software and the case study network
the control signal processing. models enabled the authors to perform a number of studies
DOLAN et al.: DISTRIBUTION POWER FLOW MANAGEMENT UTILIZING AN ONLINE OPTIMAL POWER FLOW TECHNIQUE 795
TABLE I
EXAMPLE OF BREAKPOINT AND COSTS FOR CONNECTED DG
UNITS. DG A IS DEEMED TO BE THE FIRST CONNECTED UNIT
TABLE II
GENERATOR CONNECTION PRIORITY
VII. CONCLUSION capability to calculate voltage settings and other network con-
trols to meet DNO objectives. The coordination of different ob-
The closed loop simulation results presented above demon-
jectives within an OPF formulation of ANM requires further
strates that the novel application of OPF combined with pro-
thought.
totype PFM algorithms, running on a commercially available
substation computer platform, has the potential to operate in
an online and real-time environment. OPF-LIFO solution com- REFERENCES
putation time is in the tens of milliseconds range for moder- [1] D. Roberts, Network Management Systems for Active Distribution
Networks—A Feasibility Study, DTI Distributed Generation Pro-
ately sized distribution network areas. Continuous monitoring gramme (Contractor: SP PowerSystems LTD) Contract Number:
and control leads to progressive DG set point control signals K/EL/00310/00/00, URN Number: 04/1361, 2004.
allowing for the LIFO representation to be maintained and for [2] Department of Trade and Industry, in Energy White Paper: Our Energy
Future—Creating a Low Carbon Economy, 2003 & Meeting the Energy
good resolution DG set points to be calculated in contrast to Challenge, 2007.
the DG unit capacity blocks proposed in other PFM scheme de- [3] E. M. Davidson, S. D. J. McArthur, J. R. McDonald, and P. C. Taylor,
signs. Under network outage conditions and other more onerous “An architecture for flexible and autonomous network management
systems,” in Proc. CIRED 2009, Jun. 2009.
operation conditions a solution to the circuit congestion is found [4] E. M. Davidson, S. D. J. McArthur, C. Yuen, and M. Larsson,
through DG unit constraint control but the LIFO connection “AuRA-NMS: Towards the delivery of smarter distribution networks
agreements are not honoured. The outcome in this case allows through the application of multi-agent systems technology,” in Proc.
IEEE Power and Energy Society General Meeting—Conversion and
greater energy yields across the DG units than that of conven- Delivery of Electrical Energy the 21st Century, 2008, Jul. 20–24, 2008.
tional methods, i.e., inter-tripping of all “non-firm” DG units. [5] E. M. Davidson, S. D. J. McArthur, M. J. Dolan, and J. R. McDonald,
This non-conformity of the LIFO principle of access demon- “Exploiting intelligent systems techniques with an autonomous re-
gional active network management system,” in Proc. IEEE Power and
strates the potential for OPF to be used under different connec- Energy Society General Meeting 2009.
tion principles such as DG bids for access to distribution net- [6] M. J. Dolan, E. M. Davidson, G. W. Ault, and J. R. McDonald, “Tech-
work capacity or some form of access sharing to enhance DG niques for managing power flows active distribution networks with
thermal constraints,” in Proc. CIRED 2009, Jun. 2009.
access. In summary, the novel OPF-LIFO approach has been [7] E. M. Davidson, M. J. Dolan, S. D. J. McArthur, and G. W. Ault,
found, through implementation and testing, to be viable in terms “The use of constraint programming for the autonomous management
of robustness (find a practical solution under all test cases), of power flows,” in Proc. 15th Int. Conf. Intelligent Systems Applica-
tions to Power Systems, Nov. 2009.
speed of computation and the ability to represent existing (i.e., [8] M. J. Dolan, E. M. Davidson, G. W. Ault, J. R. McDonald, F. Coffele,
LIFO) and alternative principles of access to distribution net- and I. Kockar, “Using optimal power flow for management of power
work capacity. Future work would be required to address impor- flows active distribution networks with thermal constraints,” in Proc.
44th Int. Universities’ Power Engineering Conf., Glasgow, U.K., 2009.
tant practical issues ahead of field test of the OPF based PFM [9] T. Boehme, G. P. Harrison, and A. R. Wallace, “Assessment of distribu-
method but the results presented in this paper show the potential tion network limits for non-firm connection of renewable generation,”
of the approach. IET Renew. Power Gen., vol. 4, no. 1, pp. 64–74, 2010.
[10] A. Collinson, F. Dai, A. Beddoes, and J. Crabtree, Solutions for
the Connection and Operation of Distributed Generation, DTI
VIII. FUTURE WORK Distributed Generation Programme (Contractor: EA Technology)
K/EL/00303/00/01/REP, 2003.
The analysis in this paper indicates that the performance, [11] R. A. F. Currie, G. W. Ault, C. E. T. Foote, and J. R. McDonald, “Active
power-flow management utilising operating margins for the increased
based upon reliable data, is robust since no non-convergence connection of distributed generation,” IET Gen., Transm., Distrib., vol.
situations arose. However, performance testing would require 1, no. 1, pp. 197–202, Jan. 2007.
modelling and simulating erroneous data from sensors and lack [12] European Commission, Combating Climate Change: The EU Leads
the Way. [Online]. Available: http://ec.europa.eu/publications/book-
of measurements as these would have a direct implication on the lets/move/70/index_en.htm.
network model. Misrepresentation of the actual system states [13] S. C. E. Jupe and P. C. Taylor, “Strategies for the control of multiple
would lead to erroneous control signals being sent to genera- distributed generation schemes,” in Proc. CIRED 2009, Prague, Czech
Republic, Jun. 2009.
tors and increase the probability of load flow divergence. Re- [14] European Commission, European SmartGrids Technology Plat-
cently, these issues have been discussed [24] in the context of form—Vision and Strategy for Europe’s Electricity Networks of the
voltage control and an approach to dealing with the loss of com- Future, 2006.
[15] J. Carpentier, “Optimal power flow,” Int. J. Elect. Power Energy Syst.,
munications to controllable devices presented. It is envisaged, vol. 1, pp. 3–15, 1979.
for this paper’s approach, that under these severe cases (e.g., [16] J. Carpentier, “Contribution a l’etude du dispatching economique,”
loss of or uncertain measurements or loss of controllable de- Bull. Soc. Franc. Elect., vol. 8, no. 3, pp. 431–447, 1962.
[17] H. Dommel and W. F. Tinney, “Optimal power flow solutions,” IEEE
vices attained via error checks or the communications standard) Trans. Power App. Syst., vol. PAS-87, pp. 1866–1876, 1968.
that the DNO would have to revert to traditional approaches, [18] I. Kockar and F. D. Galiana, “Combined pool/bilateral dispatch—Part
either trip all DG units or wait for protection systems to op- 2: Curtailment of firm and non-firm contracts,” IEEE Trans. Power
Syst., vol. 17, no. 4, pp. 1184–1190, Nov. 2002.
erate. This is undesirable but on the basis that active power flow [19] L. F. Ochoa, C. J. Dent, and G. P. Harrison, “Distribution network
management approaches yield higher levels of DG penetration capacity assessment: Variable DG and active networks,” IEEE Trans.
[11], [13], [25] than conventional inter-tripping schemes then Power Syst., vol. 25, no. 1, pp. 87–95, Feb. 2010.
[20] F. Capitanescu and L. Wehenkel, “Optimal power flow computations
this approach may be deemed suitable for extreme cases of error. with a limited number of controls allowed to move,” IEEE Trans. Power
In addition, further investigation of the capability of the OPF Syst., vol. 25, no. 1, pp. 586–587, Feb. 2010.
technique to represent alternative principles of DG access than [21] P. C. Taylor, T. Xu, S. McArthur, G. Ault, E. Davidson, M. Dolan, C.
Yuen, M. Larsson, D. Botting, D. Roberts, and P. Lang, “Integrating
LIFO is required to determine the technique’s feasibility and voltage control and power flow management AuRA-NMS,” in Proc.
flexibility for future connection arrangements. OPF also has the CIRED Smart Grids 2008.
DOLAN et al.: DISTRIBUTION POWER FLOW MANAGEMENT UTILIZING AN ONLINE OPTIMAL POWER FLOW TECHNIQUE 799
[22] P. C. Taylor, T. Xu, N. S. Wade, M. Prodanovic, R. Silversides, T. Ivana Kockar (S’99–M’04) received the Dipl.Ing.
Green, E. M. Davidson, and S. D. J. McArthur, “Distributed voltage (M.Sc.) degree from the University of Belgrade, Bel-
control AuRA-NMS,” in Proc. IEEE Power Engineering Society Gen- grade, Yugoslavia, and the M.Eng. and Ph.D. in 2004
eral Meeting, Jul. 2010, pp. 1–7. from McGill University, Montreal, QC, Canada.
[23] ABB. [Online]. Available: http://www.abb.com/Product/ She is a Lecturer within the Institute for Energy
seitp328/3cc7d0b2a99f2f40c1257188002facd2.aspx?product- and Environment at the University of Strathclyde,
Language=us&country=00. Glasgow, U.K. Her research interests include power
[24] M. Glavic, M. Hajian, W. Rosehart, and T. Van Cutsem, “Re- system operation planning and economics of energy
ceding-horizon multi-step optimization to correct nonviable or systems.
unstable transmission voltages,” IEEE Trans. Power Syst., vol. 26, no. Dr. Kockar chairs the IEEE PES Computing and
3, pp. 1641–1650, Aug. 2011. Analytical Methods Subcommittee (CAMS) and
[25] G. W. Ault, R. A. F. Currie, and J. R. McDonald, “Active power flow co-chairs the IEEE PES Task Force on Agent-Based Modeling of Smart Grid
management solutions for maximising DG connection capacity,” in Market Operation.
Proc. IEEE Power Engineering Society General Meeting, Jul. 2006,
pp. 1–5.
Michael J. Dolan (M’09) received the B.Eng (Hons) Graham W. Ault (M’98) received the B.Eng. degree
degree in electronic and electrical engineering from in electrical and mechanical engineering and the
the University of Strathclyde, Glasgow, U.K., in Ph.D. degree in electrical power systems from the
2006. University of Strathclyde, Glasgow, U.K., in 1993
He is currently a Research Fellow at the Institute and 2000, respectively.
for Energy and Environment at the University of Since 1996, has been researching power system
Strathclyde. His research interests include: active planning and operations issues relating to distributed
network management and sustainable energy sys- energy resources in distribution systems. He is cur-
tems for developing countries. His recent work has rently a Professor in the Institute for Energy and En-
focused on techniques and tools for real-time control vironment at the University of Strathclyde.
and operation of future energy networks.
Euan M. Davidson (M’06) received the M.Eng. Stephen D. J. McArthur (M’93–SM’07) received
degree from the University of Strathclyde, Glasgow, the B.Eng. (Hons.) and Ph.D. degrees from the Uni-
U.K., in 2000. versity of Strathclyde, Glasgow, U.K., in 1992 and
He is a Senior Research Fellow at the Institute for 1996, respectively.
Energy and Environment at the University of Strath- He is Director of the Institute for Energy and Envi-
clyde. His research interests include: multi-agent ronment and Director of the British Energy Advanced
systems; model-based reasoning; and the application Diagnostics Centre at the University of Strathclyde.
of intelligent system techniques to diagnostic, power
system control, operation, and management prob-
lems. His recent work has focused on active network
management and tools, techniques, and architectures
for intelligent or smart networks