Energy and Power Engineering, 2013, 5, 1022-1025
doi:10.4236/epe.2013.54B195 Published Online July 2013 (
Application of Risk-based Flexibility Assessment
Methods to Evaluate System Expansion Plans
Baorong Zhou1, Pei Zhang2, Jian b in C h en1, Xiaoming Jin1, Kai Hou2, Ming Niu3,
Zhao Xu3, Yunkai Lei2
1Electric Power Research Institute of China Southern Power Grid, Guangzhou, China
2Tianjin University, Tianjin, China
3Department of Electrical Engineering, The Hong Kong Polytechnic University, Hong Kong SAR, China
Received April, 2013
This paper proposed a flexibility assessment approach based on and risk assessment methodology. System planners
prioritize each planning scheme with consideration of three aspects: reliability, economics, and flexibility. In the past,
there is lack of quantitativ e indices to measure flexibility of a power system. This pap er proposes applying probabilistic
risk assessment method to quantify system flexibility. The proposed approach is demonstrated to compare two trans-
mission plann ing schemes during Guangdong expansion planning process.
Keywords: Flexibility Assessment; Transmission Scheme Prioritization
1. Introduction
Development of power systems is of the foundations of
the national economy. Thus power system planning is
always one of the most important aspects of electric in-
dustry. In order to guarantee the reliability and economy
of the power supply, as well as to meet the rapidly in-
creasing demand of electric industry, power system
planners generally establish several planning schemes,
among which, only the most reliable, economical and
flexible one will be implemented. Therefore, it is great
importance for planners to apply reasonable methods to
perform assessment and prioritization for those schemes.
Regarding assessment and prioritization, there are
clear rules and effective methods currently to conduct
reliability and economy assessment. However, there is no
recognized standard or valid approach associate with the
flexibility assessment. The flexibility of planning
schemes indicates their tolerance for uncertain factors,
major of which are listed as follows:
Uncertainty of fault probability of components such
as generator, line and transformer;
Uncertainty of power generation location, capacity,
timing, and availability of new generating facilities;
Uncertainty of future demand, scope and complex-
ity of transaction;
Uncertainty of regulatory constraints and rules.
To handle these uncertainties, McCalley and his asso-
ciates have developed a risk-based method for power
system security assessment [1]. This method considers a
predefined set of contingencies, and takes into account
both probabilities and impacts of these events. The im-
pacts are measured in terms of the voltage or current vi-
olations caused by these predefined contingencies. This
innovative method has been successfully implemented in
numerous fields such as network planning, operational
planning and op eration [2-5].
Based on the risk assessment theory, EPRI, together
with some electric power corporations, has developed a
Probabilistic Risk Assessment (PRA) methodology and
further software to perform risk assessment on power
systems [6]. The PRA program computes the probabilis-
tic risk indices and displays the results clearly with forms
and images. Taking into account both probabilities and
physical impacts of contingencies, planners are able to
have a better understanding of planning schemes, as well
as identify potential failure modes. PRA methodology
and software have been implemented in a number of
electric power corporations, such as Southern Company,
American Electric Power, KCP&L, the Eastern Inter-
connection, NYPA, ERCOT and Tri-State. At present,
PRA method is being introduced into over ten utilities for
their own studies [7-9].
This paper proposes using probabilistic risk assess-
ment method to perform flexibility assessment and quan-
tify system flexibility. Quantitative indices can be used
by system planners to prioritize planning schemes. The
second part of this paper presents PRA methodology.
Copyright © 2013 SciRes. EPE
B. R. ZHOU ET AL. 1023
The third part of this paper introduces two expansion
planning schemes of Guangdong Power and study as-
sumptions. The fourth part of this paper presents flexib il-
ity assessment results using PRA methods and prioritiza-
tion based on comparison of ri sk indi ces.
2. Risk-based Flexibility Assessment Method
Flexibility assessment measures the capabilities of the
designed power system to handle future uncertainties. As
shown in Figure 1, system operators need to know where
the potential problems are located, how likely they will
happen, and how far the operating point is close to the
boundaries [7]. Risk assessment should provide detail
information of potential dangers, just like radars. The
traditional deterministic con ting ency an alysis is unab le to
catch the diverse probabilities of events that lead to po-
tential security limit violations. PRA method takes both
likelihood and impacts of each contingency into consid-
eration, and work out a single reliability index – Prob-
abilistic Risk Index (PRI). We proposed to use these
quantitative risk indices to evaluate system flexibility.
2.1. Probabilistic Risk Indices
PRI is defined as the product of an impact by a probabil-
PRIProbability Impact (1)
where the probability quantifies the likelihood of the
simulated outage configuration, and the physical impact
quantifies the severity of the situation.
2.2. Impacts
According to different types of impacts, there are four
major risk indices: IPRI,N (voltage violation), IPRI,O (am-
perage or thermal overload), IPRI,V (voltage instability)
and IPRI,L (load loss):
Voltage Violation Index IPRI,N:
V (2)
Figure 1. Operating power system is like navigating a ship
where VIMP,i is the voltage violation from the bus upper or
lower limits caused by the ith contingency. The voltage
impact is measured in terms of kV or p.u. SVN is a set of
contingencies that causes voltage violations.
Thermal Overload Index IPRI,O:
where AIMP,i is the thermal overload above the branch
thermal rating caused by the ith contingency. The ther-
mal overload impact is measured in terms of MVA or
KA. SLO is a set of contingencies that cause overload.
Voltage Instability Index IPRI,V:
where VIMP,Si is the voltage instability impact caused by
the ith critical situation. It is measured in binary format.
If a situation causes the system becoming voltage unsta-
ble, the voltage stability impact value of is equal to 1.
Otherwise, it is equal to 0. SVS is a set of contingencies
that cause voltage instability.
Load Loss Index IPRI,L:
where LIMP,Li is the to tal load loss caused b y the ith situa-
tion. The load loss impact is measured in MW. SLR is a
set of contingencies that cause load loss.
2.3. Probabilities
In a planning context, probability is a measure of the
likelihood that the po wer system will be in a given situa-
tion at a random time in the future, and is a function of
the availability of ev ery piece of equipment in the power
iUj A
probabilityu cu c
where U represents the set of unavailable components; A
represents the set of available components; =UA,
represents the complete set of system components. u(ci)
and (1-u(ci)) represent the unavailability and availability
rates of the component i, respectively.
3. Study Data and Assumption
In 2030, the total peak lo ad demands of Guangdong elec-
tric power network will be approximately 175 MW,
among which the load demands of Pearl River Delta will
account for about 74%. Outer area power transfer into
Guangdong electric power network will be about 53.3
MW accounting for about 28% of the total load demands.
86% of 53.3MW will be transferred by HVDC. There are
two expansion planning schemes for 2030 Guangdong
electric pow er network frame .
Copyright © 2013 SciRes. EPE
The first scheme succeeds the existing 500KV double
looped network structure, as shown in Figure 2. The
eastern and western parts of Guangdong electric power
network are connected by two 500kV tie-lines which
enable 2000MW power transfer. This system includes
7972 buses, 63 70 lines a n d 57 36 transformers.
The second scheme, as shown in Figure 3, plans to
build a 1000KV UHV transmission network to connect
major nuclear power plants that are located at eastern and
western parts of Guangdong province. This system con-
sists of 7886 buses, 6349 lines and 5668 transformers.
Contingency analyses of the two planning schemes
were performed using BPA software. The contingency
analysis results provide th e detailed information of phys-
ical impact of each contingency.
The unavailability rate of a line is estimated using the
following equation:
/8760uOutageFreqRepair Time
The outage frequency is estimated using the following
/Outage FreqabZZpuPerMile (8)
where a (1/year) is the constant parameter of the forced
outage frequency; b (1/year/mile) is the proportional pa-
rameter of the forced outage frequency; ZpuPerMile
(pu/mile) is the average impedance (p.u.) per mile used
to estimate the line length; RepairTime (hour) is the av-
erage repair time (hour) after a forced outage.
ng grid
ng grid
Western gridEastern grid
Figure 2. High Voltage DC Transmission Scheme.
Figure 3. Ultra high voltage DC transmission scheme.
4. Flexibility Assessment and Prioritization
of Planning Schemes Using PRA Method
We performed risk-based flexibility assessment for two
planning schemes and calculate quantitative risk indices,
which are used as flexibility indices of planning schemes.
Prioritization can be performed by system planners based
on these indices.
Table 1 provides a summary of risk indices for two
planning schemes. With consideration of four risk factors:
overload, voltage and voltage instability, and load loss,
it clearly shows that Scheme A has lower risk comparing
with Scheme B. This indicates Scheme A has a better
flexibility than Scheme B.
Figure 4 show the risk analysis of bus voltage viola-
tions, where it is shown that the risk indices of schemes
A and B are 2.623E-5 and 1.799E-3 respectively. Obvi-
ously, the risk of scheme B is larger than that of scheme
A more than 60 times from a voltage violation risk per-
Figure 5 shows the risk analysis of transmission line
overloading, where it is shown that the risk indices of
schemes A and B are 9.778 and 10.92 8 respectively. Ob-
viously, the risk of scheme A is larger than that of
scheme B from the overload perspective, because the
overloading risk of scheme B is 1.12 times of scheme A.
Figure 6 shows the risk analysis of voltage instability,
where it is shown that the risk indices of schemes A and
B are 7.379E-3 and 3.014E-2 respectively. From voltage
instability perspective, scheme A has higher risk of volt-
age instability than Scheme B.
Table 1. Overall Risk-baesed Felxibility Analysis ResulT.
violation Overload Voltage
instability Load loss
Scheme A2.623E-5 9.778 7.379E-3 0
Scheme B1.799E-3 10.928 3.014E-2 0
Figure 4. Risk analysis of bus voltage violations.
Figure 5. Risk Analysis of line overloading.
Copyright © 2013 SciRes. EPE
Copyright © 2013 SciRes. EPE
[2] H. Wan, J. D. McCalley and V. Vittal, “Risk Based
Voltage Security Assessment,” Ieee Transactions on
Power Systems, Vol. 15, 2000, pp. 1247-1254.
[3] Y. J. Dai, J. D. McCalley, N. Abi-Samra and V. Vittal,
“Annual Risk Assessment for Overload Security,” Ieee
Transactions on Power Systems, Vol. 16, 2001, pp.
616-623. doi:10.1109/59.962405
Figure 6. Risk analysis of voltage instability. [4] W. H. Fu, J. D. McCalley and V. Vittal, “Risk Assess-
ment for Transformer Loading,” Ieee Transactions on
Power Systems, Vol. 16, 2001, pp. 346-353.
Considering the results above, scheme A obviously
outperforms scheme B in all three aspects. Therefore it
can be concluded that according to risk-based flexibility
analysis, scheme A is a better flexibility and is recom-
mended to be adopted b y the utility.
[5] M. Ni, J. D. McCalley, V. Vittal and T. Tayyib, “Online
Risk-based Security Assessment,” Ieee Transactions on
Power Systems, Vol. 18, 2003, pp. 258-265.
5. Conclusions [6] Z. Pei, S. T. Lee and D. Sobajic, “Moving Toward Prob-
abilistic Reliability Assessment Methods,” in Probabilis-
tic Methods Applied to Power Systems, 2004 Interna-
tional Conference on, 2004, pp. 906-913.
This paper introduced a new risk-based flexibility as-
sessment and prioritization method for planning schemes.
Two planning schemes of Guangdong electric power
network were studied and the risk-based flexibility indi-
ces were computed, including voltage violation index,
thermal overload index, voltage instability index and
load loss index. With this flexibility information, system
planners can evaluate the flexibility and prioritize those
project schemes. The flexibility assessment helps system
planners evaluate the operational risks when designing
the future power system, therefore builds a linkage be-
tween system planning and system operation.
[7] Z. Pei, M. Liang, L. Hopkins and B. Fardanesh, “Utility
Experience Performing Probabilistic Risk Assessment for
Operational Planning,” in Intelligent Systems Applica-
tions to Power Systems, 2007. ISAP 2007. International
Conference on, 2007, pp. 1-6.
[8] Z. Pei, M. Graham and D. Ramsay, “Prioritization of
Transmission Projects using EPRI Probabilistic Risk As-
sessment Program,” in Power and Energy Engineering
Conference, 2009. APPEEC 2009. Asia-Pacific, 2009, pp.
[9] Z. Pei, L. Shanshan, X. Ruilin, L. Xinyu and F. Li, “As-
sessing System Risk and Integrating Operation and Plan-
ning Functions for Chongqing Power using EPRI Prob-
abilistic Risk Assessment Program,” in Power and En-
ergy Engineering Conference (APPEEC), 2010
Asia-Pacific, 2010, pp. 1-6.
[1] J. D. McCalley, V. Vittal and N. Abi-Samra, “An Over-
view of Risk Based Security Assessment,” in Power En-
gineering Society Summer Meeting, 1999. IEEE, Vol. 1,
1999, pp. 173-178.