The Southern African Power Pool (SAPP) steady state security assessment using contigency analysis.

This thesis assesses the severity of potential disruptions that may occur in the SAPP (Southern African Power Pool) network by performing a contingency classification and then verifying these results by performing contingency analysis in DIgSILENT PowerFactory. The author would like to thank the following, which greatly contributed to the completion of this thesis;.

Overview
Problem Formulation
Aims and Objectives
Motivation
Hypothesis
Research questions
Methodology
Contributions
Publications in Journal and Conference proceedings

While most utilities are operational (OP), some are declared non-operational (NP) and others are merely observers (OB). Performing a contingency analysis on the SAPP network model using DIgSILENT PowerFactory to determine the security status of the SAPP network.

Overview
Power System Reliability and Security
Contingency Analysis
Contingency Ranking and Selection
N-1 Criterion

Voltage Violations
Line MVA Limit / Thermal Overload Violation

FACTS (Flexible Alternating Current Transmission System) that can be used to increase controllability and increase power transfer capability in the grid. Both of these can cause changes in transmission system voltages and power flow.

Figure 2 - 1: Contingency analysis flow chart [31]

SAPP Overview
SAPP Interconnections
SAPP Planning Criteria

The Planning Process
Transmission Planning Criteria
Transmission Planning Concepts
Planning Criteria
Power Transfer Capability
Contingency Criteria for Long Term Planning Purposes

Integration of Power Stations

Power Stations of Less Than 1000MW
Power stations of More Than 1000 MW

Voltage Limits and Targets
Equipment Loading Limits

Transmission Lines
Transformers
Series Capacitors
Shunt Reactive Compensation

Shunt Reactive Device Switching
Circuit Breakers

When all connecting lines are in operation, it should be possible to transmit the total power from the power plant to the system for each system load condition. When one connecting line is out of service (N-1), it shall be possible to transmit the total power from the power plant to the system for each system load condition.

Figure 3- 1: South African reserve margin (2002 - 2009)

Network Building and Design

ESKOM A consists of the Aries and Aggeneis power stations, ESKOM B consists of the Spitskop power station and ESKOM C the Matimba, Arnot, Camden and Normandie power stations. The sections are simply made to improve visibility during presentation taking into account the large scale of the ESKOM network and in no way reflect any actual fragmentation or division by ESKOM. 34 | Page 4-12 shows the SAPP network after all the substations in the region are interconnected to create one megastation.

Figure 4- 1: Split sections of station connectivity

Base Case Load Flow Analysis

This is the pre-contingency state of the model, before any failure or disturbance occurs on any component. 40 | Page Figure 4-18 shows that the collector of the ZESCO network before extraordinary events remains within the acceptable limit. 41 | Page Figure 4-19 shows the ZESA network before unforeseen events with all but one of the busbars within acceptable safety limits.

42 | P a g e The ESKOM A network in figure 4-20 shows that all bands fall within the required limits before the contingency.

Figure 4- 13: Upper, lower voltage limits and thermal loading legend

Contingency Ranking

Line Outage Contingency Ranking
Generator Outage Contingency Ranking

Although the results obtained from the performance index do not tell us exactly which elements violated the thermal loads and voltage limitations after the contingency, it nevertheless allows to prioritize the most severe cases which can save a lot of time for engineers. in the planning and control phase of each network. From the results shown in fig.4-26, we can find that the generator interruption that will cause the most violations are those connected to G.5, and the lowest is G.4. A contingency analysis will be performed for the top six cases to look in detail at the elements that violate the thermal and post-contingent voltage limits on both the line and the generator.

Table 4- 4: Line outage performance index

Contingency ranking

56 | P a g e The overall purpose of contingency analysis is to discover areas of weakness in the network that can be strengthened by methods including increasing transmission capacity and allows for improvements in the transmission system that will enable the network to withstand disturbances. Charts include SAPP model network chart, substation busbar chart, voltage violation chart, and thermal violation chart.

SAPP Model Network Graph

Voltage Violation

Thermal violations

Scenario 1: Contingency C.L.12 Harib – Kokerboom 2

58 | P a g e In comparison with the NAMPOWER load flow graph in fig.4-22, we can see from fig.5-2 which is under standby C.L.12 that the Kokerboom 400 Bus1 busbar rises above the safe voltage limit at the color change from blue to red. The Kokerboom 220 bus experienced a voltage drop below the safe limit and is indicated by a blue rail. The voltage base is specified with a value per unit of 0.928, which is below the required safety limits, and the after-standby voltage increases to 1.124 p.u., which is above the required safety limit of 1.05 p.u.

Figure 5- 2: NAMPOWER substation network post contingency C.L.12

Scenario 2: Contingency C.L.10 Gabarone South – Spitskop 132kV

61 | P a g e Fig.5-5 on the other hand shows that all the busbars in the network experience the voltage falling below the safe limit when under the contingent C.L.10. This change in voltage is represented by a differentiation in color as the voltage drops from a voltage that is within the required safety limits to one that is not. The bar on the graph is multi-coloured, from green on the far right to indicate that the voltage is within the specified safety limit to a dark red on the far left to indicate that the voltage now falls outside the required limit of security.

The column corresponding to the Phokojoe 220 BB is colored bright red on the far right, indicating that its base voltage of 0.877 p.u is not within certain limits and drops further to 0.833 p.u, which is represented by the dark red color on the far left of the column.

Figure 5- 5: BPC substation network post contingency C.L.10

Scenario 3: Contingency C.L.11 Harib – kokerboom 1

64 | Page Similar to the previous example, Figure 5-8 shows the kokerboom 400 Bus1 rising above the safe voltage limit with a color change from blue to red, while the Kokerboom 220 Bus experienced a voltage drop below the safe limit and is shown with a blue bus. 65 | Page In Figure 5-9, the base voltage is shown at a per-unit value of 0.93, which is below the required safety limits, and the contingency voltage rises to 1.12 p.u., which is above the required safety limit of 1.05 p.u.

Scenario 4: Contingency C.L.7 (Camden - Normandie 400_1)

67 | P a g e Figure 5-11 shows the transmission line assumed to undergo a fault, however, no significant changes are seen in the figure compared to the pre-emergency state of the network in Figure 4-22. However, it should be noted that Figure 5-11 has been enlarged to clearly show and focus on the area where the emergency occurred. From the figure we see that the Arnot Gen T6 transformer has a base load of 135% and step up to 176.9%, both of which are beyond the safety limit and therefore represented by a bar that is bright red and dark red to indicate the increase in the violation.

In the same figure, the Arnot Gen T3 transformer having a load base case of 88.7% is represented by a dark green color in the beam and drops to 84.4% after standby and is represented by a light green color as they are both in the specified safety limit.

Figure 5- 11: ESKOM C substation post contingency C.L.7

Scenario 5: Contingency C.L.13 (Infulene – Matole 275_1)

70 | Page Figure 5-14 shows that there were no significant changes in the EDM substation after the emergency, but indicates the transmission lines connected to other substations where voltage drops and increases were experienced as explained in Figure 5-13.

Figure 5- 14: EDM substation network post contingency C.L13

Scenario 6: Contingency C.L.18 (Maputo – Matola 275_1)

73 | Page Figure 5-17 does not show significant changes in emergency conditions at the EDM substation, but it does show transmission lines interconnected with other substations that experienced voltage dips and surges. Only one thermal violation was recorded and that was in the Camden Gen 8T transformer showing an increase in thermal load from 79.98% to 112.03%. This is represented in the line by a line that is colored green and turns red to indicate that a thermal overload has occurred.

Figure 5- 17: EDM substation network post contingency C.L.18

Scenario 7: Contingency C.G.5 (Synchronous Machine (2))

76 | P a g e Figure 5-20 shows that after the contingency, there is a voltage drop below 0.9 p.u on all the busbars of the BPC network. 77 | P a g e Figure 5-21 shows voltage drops on Segoditshane 200 Bus 1, Segoditshane 200 Bus 2, Gabarone No2 132 Bus 1 and Segoditshane 132 Bus 1, all of which were within the safe limit before the contingency and dropped below. contingency.

Figure 5- 20: BPC substation network post contingency C.G.5

Scenario 8: Contingency C.G.3 (Arnot Gen6)

The results show the network's response to a Gen 3 generator failure. To improve the safety of the SAPP network after the C.L.10 emergency, a shunt capacitor with a capacity of 50 Mvar was placed in the BPC network to inject reactive power into the network to increase the voltage profile. 93 | P a g e Figure 6-2 shows the SAPP network after strategic placement of the shunt capacitor to increase the security of the network.

95 | P a g e Table 6-1 shows the comparison of the voltage value after the C.L.10 contingency occurred and the results after the shunt capacitor and SVC were placed in the network to increase the network security.

Figure 5- 23: EDM substation network post contingency C.G.3

Scenario 9:Contingency C.G.9 (Camden Gen 8), C.G.7(Camden Gen 6), C.G.8(Camden

Scenario 10: Contingency C.G.10 (Gen 3)

Although not part of the top six contingencies, these results are positioned to reflect a broader range of results obtained after the contingency simulations were run. 88 | P a g e Table 5-2 collectively shows the various elements that were affected after each named contingency was executed.

Figure 5- 30: ESKOM C substation post contingency C.G.10

Security Enhancement of the SAPP Network Post Contingency C.L.10 (Gaborone south –

94 | P a g e Another analysis was carried out in which an SVC of the same rating (50Mvar) was placed in the same position as the shunt capacitor. It can be seen from table 6-1 that the results obtained after security improvement with both the shunt capacitor and the SVC were similar. The table shows an improvement in network security by 23.6% using the shunt capacitor and by 22.87% using the SVC.

Despite its slightly higher level of safety, the shunt capacitor is a more desirable option because it is more economical than most FACTS devices, including the SVC.

Figure 6- 1: BPC substation after implementation of security enhancement using shunt capacitor post contingency C.L.10 (Gaborone south – spitskop)

Security Enhancement of the SAPP Network Post Contingency C.L.12 (Harib-Kokerboom2)

Most substations are restored to the correct voltage value within accepted safe limits as indicated by the blue and green colors. 99 | Page Table 6-3 shows the busbar voltage value before and after kokerboom 2 compensation using a shunt reactor. Before compensation, the busbar voltage was above the required limit of 1.05p.u, but dropped to 0.95p.u, which is within the safety limit.

Figure 6- 4: NAMPOWER substation after security enhancement implementation post contingency C.L.12 (Harib - Kokerboom 2)

Conclusion

Future work

Kelechi, Aqequacy analysis and security reliability evaluation of bulk power systems, Journal of Computer Engineering, vol.11, issue 2, pp 26 – 35, May-June 2013. Swarup, “Static security assessment and evaluation of power systems using remote system equivalents,” International Journal of electrical power and energy systems, vol. Davidson, “Energy planning in a smart grid environment - A case study of South Africa,” in Proceedings of the IEEE power engineering society (PES) 2013 Meeting, Vancouver, BC, Canada, 2013.

Söder, “Improving power system security by optimal placement of UPFC,” in 4th IASTED Asian Conference on Power and Energy Systems, AsiaPES 2010.