Analysis of the Use of Universal Distribution Factors in SEC Power Grid

Research Inventy: International Journal of Engineering And Science
Vol.6, Issue 4 (April 2016), PP -58-64
Issn (e): 2278-4721, Issn (p):2319-6483, www.researchinventy.com
58
Analysis of the Use of Universal Distribution Factors in SEC
Power Grid
A. Alodhaiani 1,*
, Y. Alturki 2
, M. El-Kady 2
1
Saudi Electricity Company (SEC), Riyadh, Saudi Arabia
2
Department of Electrical Engineering, SEC Chair for Reliability and Security, King Saud University, Riyadh,
Saudi Arabia
Abstract: Distribution factors have been extensively used in many power system analysis and planning studies.
In recent power system studies, the AC distribution factors are insensitive to the operating point and relatively
sensitive at certain degree to changes in network topology. These factors are linear approximations of
sensitivities of variables with various inputs. This paper presents the calculation of the universal distribution
factors (UDF’s) applies them on several practical scenarios of Saudi Electricity Company (SEC) power grid.
The results are analyzed and evaluated considering various system conditions of SEC load. The results show
that the accuracy of the used approach is acceptable compared with exact method. This is practically beneficial
to SEC in computing its grid complex power flows using UDF's at the base case without the need to recalculate
UDF’s which save efforts and time.
Key words: Distribution Factors, Sensitivity Factors,Power System, Electricity Market.
I. Introduction
AC distribution factors and its application play an important role in operation and control of power
systems and have been addressed extensively in the literature. In this respect, considerable research efforts have
been expended during recent years and it’s going on in developing theories, new techniques, policies and
required criteria for distribution factors and its application in operation and control of power system. A novel
model for evaluating universal distribution factors that is suitable for an extensive range of power system
pricing studies and power system planning and analysis were introduced in [1]. The author present the
computation of the universal distribution factors in addition to their sensitivities in with respect to the voltage
profile of line. The availability and utilization of the existing transmission systems and the operation of the
competitive markets in electricity system are very much affected by the congestion. The distribution factors are
linear approximations of sensitivities of variables with respect to various inputs and are computed for a specified
network topology and parameter values. Hence, they are very important in designing the congestion model for
various market applications. The authors in reference [2] investigated the analytical characteristics, the
robustness and the quality of the approximations provided the power transfer distribution factors (PTDFs).
These factors have provided a reliable approximation for large power system networks. The errors of the
approximations are within the permissible for a broad range of conditions including contingencies.The concepts
and the development of the AC distribution factors in power systems has been addressed in the literature [3-5].
In this regard, the authors of reference [3] presented a solution algorithm to improve the poor accuracy of
generalized generation shift distribution factor (GGDF). A modified power transfer distribution factor (MPTDF)
was used to consider the nonconformity of the demand change. When the system operating point is shifted to a
new one, the power flows in the line can be got directly without running a load flow program. By means of four
example systems, the potential behavior of the proposed method was demonstrated with very fast execution time
and with high accuracy compared with the fast-decoupled load flow (FDLF) technique. On the other hand, the
authors of reference [4] has analyzed the distribution of source power from the energy source, by deriving and
defining the complex power dynamics distribution factors based on the circuit theory. The form of these factors
shows that the power distribution is not only related to the network topology and circuit parameters, but also
related to the dynamic state of the system. The source powers are linear distribution in the network according to
these factors and whereas the branch flows and losses in the network are different forms of source powers. The
example demonstrated that the results calculated by proposed method are unique and they satisfy complex
power conservation. By using the Distribution factors, the contributions of the transmission users towards line
flows in an electricity market are identified and the transmission charges are fairly fixed. However, due to some
limitations, the Distribution factors cannot be applied for the wider applications of electricity market practices.
In reference [5], a new technique was proposed to evaluate the transmission line flows more accurately. The
new justified distribution factors approach has shown better result than other commonly used distribution factors
when tested in a realistic electricity.

Analysis of the Use of Universal Distribution…
59
Power transfer distribution factors depend on the operating point and topology of an electric power system.
However, in recent researches, the power transfer distribution factors were shown theoretically to be relatively
insensitive to the operating point. The authors of reference [6] provided empirical corroboration of this
theoretical result. The several researches and test conducted on the three principal interconnections namely in
North America, the Eastern interconnection, the Western interconnection, and the Electric Reliability Council of
Texas (ERCOT) interconnection has shown consistent observations. On the other hand, the authors of reference
[7], have studied the Power Transfer Distribution Factors (PTDF) and the underlying flow-based model
proposed for capacity determination and allocation in electricity grids. Several factors such as topological and
seasonal changes as well as zone building which influence PTDF coefficients are critically analyzed through AC
power flow simulation of the Union for the Co-ordination of Transmission of Electricity (UCTE) transmission
network (transmission system operators in continental Europe). Finding of border capacities are also analyzed.
Also, the author of reference [8] proposed a methodology to create a family of zonal Power Transfer
Distribution Factor (PTDF) matrices which can be used to find the power flows accurately for the Zonal market
under dynamic condition. Introduction of data mining methods for selected typical hours based on area balance
and PTDF matrices of these typical hours calculated to form a family are tested on the New England system. It
is observed that the error, when using a family of PTDF matrices for estimating cross-border power flows in a
zonal market model is very much less than the method of using only one PTDF matrix.
A model for calculatingdistribution factorswere applied on a several scenarios of SEC power grid in this
paper. We analyzed the calculationof UDF’s on the SEC grid at all scenarios consideringvarious percentages of
the SEC load. Theevaluation of the UDF’s with respect the operating point and network topology were
alsodiscussed.
II. Universal Distribution Factors (UDF’s)
1.1. UDF’s formula
𝑁𝑏 = 𝑁 + 𝑁𝑑 where,𝑁𝑏 is the number of the buses in the power system with N and 𝑁𝑑 is the number of the
generation buses and load busesin the power systemrespectively. Also, 𝑁 𝑚 is the number of the transmission
branches. We can calculate line complex power flows (𝑆𝑝𝑞 ) between bus 𝑝 and bus 𝑞from follow equation;
𝑆𝑝𝑞
∗
= 𝑉𝑝
∗
𝐼𝑝𝑞 (1)
Where𝑆𝑝𝑞 , 𝑉𝑝 and 𝐼𝑝𝑞 denote theline complex power flows of 𝑝𝑞line at bus 𝑝, voltage at node 𝑝, and 𝑝= 1, 2, …
, 𝑁 and current in line 𝑝𝑞, respectively.
The complex power flow between buses 𝑝 and 𝑝on line 𝑚can be expressed as;
𝑆𝑝
∗
= 𝑙 𝑉𝑝
∗
+ 1 − 𝑙 𝑉𝑞
∗
𝐼 𝑚 = 𝜑 𝑚 𝐼 𝑚 (2)
Where 𝑙denote the location on line 𝑚at which𝑆 𝑚 is evaluated. For illustration, if 𝑙 = 0this illustrate that𝑆 𝑚 is
evaluated at bus 𝑞, if 𝑙 = 1this illustrate that 𝑆 𝑚 is evaluated at bus 𝑝 and if 𝑙 = 0.5this illustrate that 𝑆 𝑚 is
evaluated at the mid-point on line 𝑚, that is;
𝑆𝑝
∗
= 𝑉𝑝
∗
+ 𝑉𝑞
∗
/2 𝐼 𝑚 (3)
Where; 𝐼 𝑚 is the current on line m.
The bus admittance matrix 𝑌 of the grid can be determined as follow;
𝑉 = 𝑍𝐼 or 𝑉 = 𝑌−1
𝐼 (4)
Diagonal matrix 𝜑 can be written as;
𝜑 𝑚 = 𝑑𝑖𝑎𝑔𝑜𝑛𝑎𝑙 𝜑1, 𝜑2, … , 𝜑 𝑛𝑚
The 𝑛 𝑚 -vector of complex conjugate linepowers can be written as follows;
𝑆 𝑚
∗
= 𝜑 𝑚 𝐼 𝑚 (5)
Where 𝑆 𝑚
∗
and 𝐼 𝑚 denotes the 𝑛 𝑚 vector of complex conjugate linepowers {𝑆 𝑚 } and 𝑛 𝑚 vector of complex line
currents {𝐼 𝑚 }, respectively.
The bus incidence matrix that is A branch-to-node incidence matrix where 𝐴 = ( 𝑛 𝑚 × 𝑛 𝑏 ), which consist of
one row for each branch and one column for each bus with an entry 0, 1 or -1. Where,𝐴 𝑏𝑚 = 0 if the branch 𝑚
is not connected to the bus 𝑏, 𝐴 𝑏𝑚 = −1 if the current in branch 𝑚 is directed toward bus 𝑏 and𝐴 𝑏𝑚 = 1if the
current in branch 𝑚 is directed away from bus 𝑏.
Therefore;

60
𝑉𝑚 = 𝐴 𝑉 (6)
Hence,
𝐼 𝑚 = 𝑌 𝑃
𝑉𝑚 = 𝑌 𝑃
𝐴 ( 𝑌−1
𝐼 ) (7)
Where, 𝑌 𝑃
primitive admittance matrix of two coupled branches. The diagonal elements of the primitive
admittance matrix represent line self-admittances and off-diagonal elements represent mutual line admittances.
Hence;
𝜑 𝑚 𝐼 𝑚 = 𝜑 𝑚 𝑌 𝑃
𝐴 𝑌−1
𝐼 (8)
The bus voltage matrix can be written as;
𝐸 = 𝑑𝑖𝑎𝑔𝑜𝑛𝑎𝑙 𝑉1, 𝑉2, … , 𝑉𝑛
Where 𝐸 is the diagonal matrix of bus voltages. The n-vector 𝑆∗
of complex conjugate bus powers can be written
as follows;
𝑆∗
= 𝐸∗
𝐼 (9)
or
𝐼 = 𝐸∗−1
𝑆∗
(10)
Therefore;
𝑆 𝑚
∗
= 𝜑 𝑚 𝐼 𝑚 = [ 𝜑 𝑇 𝑌 𝑃
𝐴 𝑌−1
) 𝐸∗ −1
𝑆∗
= 𝑈𝐷𝐹 𝑆∗
(11)
The universal distribution factors matrix 𝑈𝐷𝐹 relates line complex power flows 𝑆 𝑚 to the bus injected complex
powers𝑆𝑏 , with 𝑛 𝑏 and 𝑛 𝑚 denoting, respectively, number of buses and number of lines in the system, as
follows;
𝑆 𝑚 = 𝑈𝐷𝐹 𝑆 (12)
Where is the universal distribution factors matrix UDF is;
𝑈𝐷𝐹 = 𝜑 𝑚
∗
𝑌 𝑃∗
𝐴 𝑌∗−1
𝐸−1
(13)
1.2. UDF’s Evaluation
To evaluate the validity and the effectiveness of the computedUDF’s, the system is analyzed and studied with
applying a number of scenarios. These scenarios includes a change demands and generation. Line complex
power flows are calculated as in equation (14) in different scenarios.
𝑃𝐿 𝑛𝑒𝑤 = 𝑈𝐷𝐹 𝑃𝐵 𝑛𝑒𝑤 (14)
Where 𝑷 𝑳 𝒏𝒆𝒘 is the new line complex power flow after change the bus injected power (𝑷 𝑩 𝒏𝒆𝒘),
𝑃𝐿 𝑛𝑒𝑤 = 𝑃𝐿0
+ ∆𝑃𝐿 (15)
and
𝑃𝐵 𝑛𝑒𝑤 = 𝑃𝐵0
+ ∆𝑃𝐵 (16)
Which ∆𝑃𝐿 and ∆𝑃𝐵denotes the changes in the lines complex power flows and the changes in bus injected
powers respectively.
II. Case Studies
In this section, the universal distribution factors (UDF’s) model proposed in [1] is applied onthe 2012 real
system of Saudi Electricity Company (SEC). SEC grid consists of 2592 buses, 711 generation units, 2251
transmission lines, 1898 transformers, 35 line shunts, 248 switched shunts and 139 tie-lines. Figure 1 shows the
consumption of electricity in Kingdom of Saudi Arabia (KSA) which is totally covered by SEC interconnected
network (generation, transmission and distribution). The electricity peak load of KSA in SEC power grid
happened during the summer season at48138 MW which was recorded in 21 July 2012.

61
Fig. 1 Consumption of the electricity in KSA
The real system of Saudi Electricity Companygrid is reduced to 15 tie lines and9 buses as shown in Figure 2.
The9 buses of the reducedmodel denotes four generations owned by SEC, four load areas and one independent
power plantowned by Dhurma Electricity Company. Thefour load areas represents central operating area
(COA),west operating area (WOA), east operating area (EOA) and the south operating are (SOA).The 15 tie
lines representthe corridorsconnected all loads and generators. The UDF’s have been determined taking in to
account all the several scenarios applied on the reduced model of real SEC system.
Fig. 2Reduced model of SEC interconnected system
The UDF’s are calculated at the base case, which is the solved base case of 2012 SEC network.Threepractical
simulated scenarios were conducted to examine the validity and accuracy of using the universal distribution
factors of the base case when operating points and network topologies of SEC electric power system change.
The results are analyzed and discussed.
Table 1shows the details of the three scenarios that have been conducted to evaluate the application of universal
distribution factors on SEC network.
Table 1. Details of the simulated scenarıos for the UDF’s evaluatıons
Scenarıo no. Scenarıo Descrıptıon
Base case Solved base case of 2012 with no contingencies
Scenario no. 1
10% lossof generation owned by SEC generation
Company SEC – GEN (D) bus No.811330
Scenario no. 2 10% Increase of demand WOA bus#820001
Scenario no. 3
25% Outage of line between SEC-GEN (A) bus
No.820002 and COA demand bus No. 818003
0
5,000
10,000
15,000
20,000
25,000
30,000
35,000
40,000
45,000
50,000
KSA2012Load(MW)
Time (Hours)
2012 KSA Load Duration Curve (Hourly)
48,138 MW, 21/07/2012
1 1000 2000 3000 4000 5000 6000 7000 8000 8784

62
IV. Results and Discussion
4.1 UDF’s analysis at base case
The universal distribution factors matrix isdetermined at the base case with no contingency cases at parameter 𝒍
= 0.5 (𝑺 𝒎 is evaluated based on voltages at the mid-point of line m). The calculated universal distribution
factors of all reduced network lines are shown in Table 2.
From line no. 1 in Table 2, it is interesting to notice that the bus no. 811357 has the largest DF value in this line,
while the other buses have less influence on this line. In other words, bus 811357 is generation bus and the
amount of the transmitted power through this line is very high. DF’s values of the buses 844004, 830001 and
840011 are much smaller than the rest of the buses, which indicates that the amount of power injection or
withdrawal change in these buses have less influence on the tie-line no. 1 compared with the other buses. From
line no. 12 in Table 2, we notice that the bus no. 813257 has DF value while the other buses do not have DF
values because of this bus is connected only to one bus in the reduced network and the number of this bus is
818003. The rest lines have the same properties as line no.1.
The UDF’s matrix can be used to measure the sensitivity of the power flow to the power variation from the buses.
Therefore, we can utilize the UDF’s matrix to indicate the load demand in the system. The real and imaginary values of the
computed UDF’s have positive and negative values as indicated in Appendix A. These values are leading us to two points,
the first one is; if the UDF’s values positive, then the power flow on the corresponding branch is decreasing, and the second
point is; if the UDF’s values negative, then the power flow on the corresponding branch is increasing.
4.2 UDF’s evaluation at all scenarios
The line power flows have been determined by using one universal distribution factors matrix calculated at the
base case. These line power flows have been calculated at all the scenarios as shown in Table 3 considering
contingency cases as represented in Table 1 and using the UDF’s at the base case. The calculatedline power
flows are compared with the exact line MVA power flows obtained by AC power flow at each line. Table
3shows the error values between the exact and computed line MVA power flows at all scenarios.

63
Results showthat the absolute value of error percentages between exact and calculated line power flows for
all the network lines at scenarios no. 1 and 2 are small and negligible. It is observed that the error values for
scenario no. 1 and 2 ranges from 0% to 2.7%. Therefore, the computed UDF’s at the base case can be used on
SEC network to calculate the line complex power flows for all the network lines. It can be concluded that when
system load increases up to 10% or the network generation decreases to the percentage of 10% reduction, SEC
can use UDF’s of the base case with acceptable accuracy. This is useful to SEC as it saves time and effort by
avoiding calculating UDF’s and AC powers flows for network changes with these ranges.
On the other hand, simulated results show that the absolute value of error percentages for all the network
lines at scenario no.3 are slightly high for some lines and almost areequal to the percentage of the corridors
outages for some lines.The error values of this scenario ranges from 0% to about 10% for all lines except for
line 11 at which the error reaches 19.36%. . Therefore, the computed UDF’s at the base case can be used at
certain degree on SEC network to calculate the line complex power flows for some of the network lines, which
gives reasonable results except on line no. 11.
It is worth noticing that error percentages for scenarios 1 and 2 are much lower than those of scenario 3. So,
one can conclude that change of network topology may lead to moreeffect on UDF’s resulting in much higher
errors than the situation when the change is in system load or generation.. The universal distribution factors are
more sensitive at certain degree to the network topology than to the operating point.
V. Conclusion
This paper presented the universal distribution factors UDF’s and it applied them on different practical scenarios
on SEC power grid. The evaluation of the UDF’s on SEC network considering several scenarios including
different variations of the demands as well as changes in network topology conducted but due to limited space
three scenarios were presented. The UDF's at base case can only be usedon a changed network if the difference
between their results and the actual calculated line power flows of the new changed network is negligible (i.e.
acceptable error), otherwise, UDF's should be recalculated.
In this paper, three scenarios including generation and system demands changes as well as outage of a tie-line
of SEC power grid were simulated. Comparison of simulated results between exact and calculated line power
flows for all the network lines shows that the absolute value of the error percentages are acceptable for demand
or generation changes as represented in Table. 1. While for system topology change such as line outage,the
errors are slightly high for some lines and almost equal to the percentage of the corridors for some lines.In
conclusion, the calculated UDF’s at the base case of SEC grid can be used to calculate the line power flows for
all the network lines in several expected and critical scenarios of the SEC power network with acceptable
accuracy. This is beneficial to SEC operation as it saves time and effort on real time when such contingencies
occur on its real network.

64
VI. Acknowledgment
This work was supported by Saudi Electricity Company (SEC) through its Chair in Power System Reliability
and Security at Department of Electrical Engineering, King Saud University.
References
[1] M. A. El-Kady, "Universal Distribution Factors for Power System Open-Market Studies", IET Gener. Transm. Distrib., Vol. 5,
No. 3, 2011, pp. 356-359.
[2] M. Liu and G Gross, “Effectiveness of the Distribution Factor Approximations Used in Congestion Modeling”, Proceedings of
the 14th Power Systems Computation Conference, June 24-29, 2002, Sevilla, Spain.
[3] Y. C. Chang, W. T. Yang and C. C. Liu, “Improvement on GGDF for power system security evaluation”, IEE Proc. Gener.
Transm. Distrib., Vol. 141, No. 2, March 1994, pp. 85-88.
[4] A. Dongping, B. Hai and Y. Yihan, "A New Method of Calculating Complex Power Dynamic Distribution factors by Circuit
Theory", Proceeding of IEEE Power and Energy Engineering Conference (APPEEC), Asia-Pacific, 2010.
[5] N. H. Radzi , Z. Y. Dong and R. C. Bansal, “Justified distribution factors approach for pool modelling”, IPEC Conference
Proceedings, Singapore, Oct. 2010, pp. 117-1122.
[6] K. Dixit, R. Baldick and T. J Overbye, “Empirical Analysis of The Variation of Distribution Factors with Loading”, IEEE Power
Engineering Society General Meeting, Vol. 1, June 2005, pp. 221 – 229.
[7] C. Duthaler, M. Emery, G. Andersson and M. Kurzidem, “Analysis of The Use of PTDF in the UCTE Transmission Grid”, 2008.
[8] D. Huang, “Methodology of Dynamic PTDF for the Zonal Model”, IEEE Power Engineering Society General Meeting, Vol. 1,
June 2005, pp. 221 – 229.

Analysis of the Use of Universal Distribution Factors in SEC Power Grid

More Related Content

What's hot (19)

Similar to Analysis of the Use of Universal Distribution Factors in SEC Power Grid (20)

Recently uploaded (20)

Analysis of the Use of Universal Distribution Factors in SEC Power Grid