Energy and Power Engineering, 2013, 5, 950-953
doi:10.4236/epe.2013.54B182 Published Online July 2013 (http://www.scirp.org/journal/epe)
Method of Electricity Supply Reliability Need
Assessment for Multi-type Power Users Area
Yihao Zeng1, Lin Guan1, Ha n Wu2
1School of Electrical Power, South China University of Technology, Guangzhou, China
2Fujian Electric Power Research Institute, Fuzhou, China
Received April, 2013
ABSTRACT
With electric power outage influ ence on different type users analyzed, indexes for assessing supply reliability need level
is designed. Based on fuzzy ev aluation method, a regional power supply reliability lev el assessment method which util-
izes qualitative data is presented.
Keywords: Power Supply Reliability Need Assessment; Indexes for Need Assessment of Supply Reliability; Regional
Supply Reliability Needs
1. Introduction
Power supply enterprises value reliability management,
but need of reliability lacks of attention. Analysis and
assessment of power users reliability need is nearly blank,
so power enterprises can only obtain insufficient need
information via users’ load rank. Otherwise, for the rea-
son that projects for improving reliability cost money and
power outage causes users’ loss, it is suggested that sup-
ply companies provide customers suitable reliability ac-
cording to their requirements and charge them price of
reliability [1-3]. A regional power supply reliab ility need
assessment method which utilizes qualitative data is pre-
sented in this paper, and it may be complementary refer-
ence for distribution project feasibility analysis.
2. Power Outage Influence on Important
Users
Here is a brief analysis of outage influence on typical
users.
1) Towards high-rise residential buildings, power fail-
ure causes worst effect on high-rise ones, compared to
low-rise ones:
 Due to the existence of elevators, high-rise flats
require higher reliability. Because of a power failure
people will be inconvenient to go upstairs and
downstairs, even will be trapped in elevator. Al-
though interruption duration is short, it takes quite
long to restart an elevator by relevant administrator.
From the above, high-rise housing can endure nei-
ther long-hours power loss nor frequent interrup-
tion.
 Subject to press conditio n provided by water supply
company, departments on the sixth floor and above
have water supplied in secondary water supply
mode with pumps and tanks. After a short-time
power outage, water supply will be interrupted.
2) Towards numerous industrial customers, their load
rate is rather high and single user consumers much elec-
tric energy. Power interruption causes industrial enter-
prises’ abnormal downtime and production reduction,
meanwhile sale income of power supply company will be
affected. Otherwise, restart cycle of some industrial
equipments and production processes is rather long, like
electrolytic aluminium and Iron &steel smelting, their
owners ask for low interrupt frequency.
3) Towards commercial users, there are several sort of
load including lighting, elevator running, heating &
cooling via air conditioner, checkout. Any sort of load
listed above should not be interrupted even for a short
while, otherwise commercial operation will be severely
affected. What is more, since some heavy industrial cus-
tomers have a large proportion of load ranked level 1 and
level 2, sudden power outag e may well result in personal
injury.
4) Roads in city are divided into 4 grades, among them
main road and fast speed road are in strong position of
traffic. Once traffic lights turn blind due to power loss, it
is probable to have traffic jams and accidents.
5) Government departments take political, economic,
cultural functions and so on, so they are surly of special
importance. Therefore, government users’ supply reli-
ability needs to be guaranteed.

Y. H. ZENG ET AL. 951
3. Indexes of Need Assessment of Supply
Reliability
Towards high-rise residential building users’ needs of
reliability, HP and HC indexes are defined below:
R
S
C
HP= C (1)
H
R
C
HC= C (2)
where CR, CS, CH are annual residential electricity con-
sumption, annual total electricity consumption, annual
electricity consumption of high-rise residential building
users in specific area, respectively.
Towards commercial users, CP and CC indexes are
defined below:
C
S
C
CP= C (3)
C
C
V
CC= C (4)
where CC, VC are annual commercial electricity con-
sumption, annual commercial value of produc tion in spe-
cific area, respectively.
Towards industrial users, IP and IC indexes are de-
fined belo w:
S
C
IP= C
I
(5)
V
IC= C
I
I
(6)
where CI , VI are annual industrial electricity consump-
tion, annual industrial value of production in specific
area , respectively.
Towards traffic light users, TP and TC indexes are de-
fined belo w:
S
N
TP= CT (7)
TC=
R
S
S (8)
where NT, SR, S are number of traffic lights, area of main
roads & fast speed roads, total area in specific region,
respectively.
Towards government users, GP and GC indexes are
defined below:
S
C
GP= CG (9)
G
GC=N (10)
government departments, number of government de-
partments in specific area, respectively.
where CG , NG are annual electricity consumption of
4. Fuzzy Assessment Method of Regional
Am, indexes HP, CP, IP,
ill be divided into 5 levels (Lv.1,
Lv
Supply Reliability Needs
ong indexes listed in chapter 3
TP and GP reflect the electricity consumption scale of
different sorts of users, they are used for calculation of
weigh coefficients of corresponding sort of users and
named index set W. Indexes HC, CC, IC, TC, GC reflect
reliability need level of different sorts of users, they are
used for calculating the fuzzy assessment array and
named index set C.
All the indexes w
.2, Lv.3, Lv.4, and Lv.5) and level-related each in-
dex’s membership corresponding to every level will be
obtained based on fuzzy assessment method, where Lv.1
is top level and stands for highest reliability need.
Among index i (i=1,2,···,10), which in turn corresponds
to the index in index set W and C, set level standard val-
ue of every level for each single index, presented as bi1,
bi2, bi3, bi4, bi5 here. Membership functions selected are
listed as Equation (11)-(13) [4]. Level standard value of
every level to various indexes depends on the situation.
1
2
12
12
2
1
() , [,)
0,[0, )
ii
ii
iii ii
ii
ii
gb
rg g bb
bb
gb







(11)
1,[, )gb

(1) (1)
(1)
(1) (1)
(1) (1)
(1)
,[, ),2,3,4
()0,[, )[0,),2,3,4
,[ ,),2,3,4
iji iijij
ij ij
ij iiijij
iij iijij
iji j
bg
gbb j
bb
rgg bbj
gb gb bj
bb











 





(12)
5
4
55
45
4
1,[0, )
(),[, )
0,[, )
ii
ii
iii ii
ii
ii
gb
bg
rgg bb
bb
gb








4
(13)
where rij(g) is index i’s membership of Lv .j. Assessment
i
array R is then gained with index set C’s membership as
shown in Equation (14).
rr

1 11 12
21 222
12
n5, 5
m
m
nn nm
r
rr rm
rr r







R




(14)
Towards indexes in set W, their assessment level re-
su :
To
lts can be obtained by biggest membership principle.
Biggest membership principle is applied in this way[5]
each index i ,let its membership values of level 1 to
level 5 are m1, m2, m3, m4 and m5, respectively. If mk =
max{mj}, j= 1-5, then assessment level of index i (named

Y. H. ZENG ET AL.
952
nce-comparison method is to be
in
Third, each w
la
(16)
With weight vector K obtained above, regional reli-
ab

(17)
According to biggest membership principle,
re
+80·dA2+6 0·dA3+40·dA4+20·dA5
5. Application
es of indexes and assessment data of
h-percentage high-
ris
di) is k, namely di=k.
Then modified seque
troduced to obtain weight vector K={kHP, kCP, kIP, kTP,
kGP} from set W’s assessment level dW={dHP, dCP, dIP,
dTP, dGP}, element positions in W and K keep the same.
First to have elements in dW sorted from small to large,
let them be dW = {d1,d2,…,d5}, then K = {k1,k2,…,k5}.
Second, reference to Table 1, comparison coefficients
between indexes can be determined as Equation (15)[5]:
ki-1 / ki=ri (i=5, 4, 3, 2) (15)
eight of weight vector K can be calcu-
ted with Equation (16).
1
5 5
51
2
1 (5,4,3,2)
niii
mnm
krkrki







ility needs membership D can be calculated with Equa-
tion (17) .

12345
= dAdAdAdAdADK R
regional
liability needs assessment level can be obtained. How-
ever, in order to utilize D’s membership data[6], method
to get comprehensive assessment score G is shown in
Equation (18) .
G=100·dA1
(18)
Level standard valu
4 selected areas are listed in Table 2.
Area A is a residential area with hig
e buildings, and except high-percentage residential
electricity consumption the consumption of other types is
rather small. Except its low-percentage of high-rise build-
ings, Area B is similar to area A.
Area C is an ordinary industrial park with high-per-
centage industrial electricity consumption and there is
only little amount of consumption of other types. Area D
is a high-tech zone with the same percentage industrial
electricity consumption as area C, and its production
value per unit consumption is very much.
Weight vectors of 4 selected areas are listed in Table 3.
For area A and B, kHP is bigger than other weights. For
area C and D, kIP is rather big due to the fact that indus-
trial electricity consumption is major load of both areas.
Therefore, modified sequence-comparison method works
well and weight vectors correctly reflect features of se-
lected areas.
Table 4 contains details of reliability need level about
selected areas. Take area A for instance, membership of
Lv.1 is quite large, but membership of other level is
small. Some users require high quality of reliability and
quite a few users’ reliability need is ordinary or lower
than the average. Otherwise, the scores point out reliabil-
ity needs of selected area. Based on scores of area reli-
ability needs, reliability need sorted results of selected
areas in descending order is D>A>C>B. Such reliability
needs assessment result may be good reference for power
grid planners. Funds should be allocated to those areas
with high scores of area reliability needs.
Table 1. Reference on determining rk.
rk Description
1.0 dk-1=dk
1.5 dk-1-dk=1
2.0 dk-1-dk=2
2.5 dk-1-dk=3
3.0 dk-1-dk=4
Table 2. Level standard values of indexes and assessme nt data of 4 selected areas.
Level standard values Area assessment data
Index Lv.5 Lv.4 Lv.3 Lv.2 Lv.1 A B C D
HC [%] 5 10 20 30 40 55 20 10 10
HP [%] 5 10 12 14 20 60 60 12 12
CP [%] 1.
CC h] 140 140
IC [h]
GC .]
0. 0.1. 1.1.0.1.
23 2.46 3.07 3.68 4.30 2 2 2.5 2.5
[yuan /kW73 145 181 218 254 140 140
IP [%] 30 50 70 80 90 10 10 80 80
yuan /kW9 18 23 2 7 32 20 20 23 50
GP [%] 0.2 0.40 0.8 1.2 1.73 0.2 0.2 0.4 0.4
[1/sq.km2 3 4 5 6 4 4 4 4
TC [km/km2] 48 96 2 44 68 2 0.2 2 1.2
TP [1/sq.km.] 0.5 1 2 5 10 1 1 2 2

Y. H. ZENG ET AL. 953
Weight v resu. Table 3.ectorlts
Area
Index A B C D
kHP 0 0 0 0 .43.43.21.21
kCP 0.17 0.17 0.14 0.14
kIP 0.11 0.11 0.31 0.31
kGP 0.11 0.11 0.14 0.14
kTP 0.17 0.17 0.21 0.21
Rely neTable 4.iabilited assessment results.
Lev e l members h i p
Are Lv.5 .2 Lv.1 re a Sco
Lv.4 Lv.3 Lv
A 0.18 0.22 0.16 0.00 0.43 65.32
B 0.18 0.22 0.59 0.00 0.00 48.18
C 0.01 0.34 0.62 0.03 0.00 53.52
D 0.01 0.34 0.34 0.00 0.31 65.33
6. Conclusions
ased fuzzy assessment method for
utilizes objecti indexes to avoid suity. is me-
REFERENCES
[1] Y. T. Song, Dng, et al., “Power
Supply Reliabilityetwork of
and Future
echanism Adapted to Electric-
Indi-
cation,” China Electric
n,” 2nd Edition,
A qualitative-data-b
assessing regional supply reliability need s is presented in
this paper. This method combines modified sequence-
comparison method and its weight vector calculation
thod can help power grid planners to implement feasibil-
ity analysis.
vebjectiv Th
. X. Zhang, C.-H. Lia
Planning for Urban Power N
China Southern Power Grid,” Power System Technology,
Vol. 33, No. 8, 2009, pp. 48-54 (in Chinese).
[2] J. Zhao, C. Q. Kang, Q. Xia, et al., “Power System Reli-
ability in electricity Market Current S Tatus
Prospects,” Automation of Electric Power Systems, Vol.
28, No. 5, 2004, pp. 6-10.
[3] P. Zhou, K. G. Xie, J. Q. Zhou, et al., “Reliability Power
Price and Compensation M
ity Market Operation Environment,” Automation of Elec-
tric Power Systems, Vol. 28, No. 21, 2004, pp. 6-11.
[4] H. X. Yang, K. G. Xie, K. Cao, et al., “Reliability Elec-
tricity Price Model for Power Market Considering
cator Weight,” Power System Protection and Control,
Vol. 39, No. 16, 2011, pp. 67-73.
[5] Y. X. He, D. N. Liu, T. Lu, et al., “Power Comprehensive
Evaluation Method and Its Appli
Power Press, Beijing, 2011, pp. 42-45.
[6] D. Du, Q. H. Pang and Y. Wu, “Modern Comprehensive
Evaluation Method and Case Selectio
Tsinghua University Press, Beijing, 2008, pp.40.