Identification of system vulnerabilities in the Ethiopian electric power system

1 Introduction

The past decade has witnessed a big growth in Ethiopia’s electricity sector. Between 2012 and 2016,the number of electricity customers of EEP reached over four million from two million at the beginning of the period [1]. This steep growth in the electricity sector has led to significant constraints in EEP. Moreover, while the generation has been at the forefront of energy developments in Ethiopia, the weaknesses in the transmission network are not receiving their due attention. These weaknesses affect the customers, the government, the economy and social welfare of the society. Due to these facts, fault(s) in the transmission system are threatening the operation of the power system, and thereby the national grid system’s vulnerability is increased from time to time.

The main HV levels of the power transmission systems are 400 kV, 230 kV, and 132 kV. The 500 kV transmission system is on the way to come. Among the voltage levels, the 400 kV and 230 kV are the most important and are responsible for the bulk power flows among the eight regional power systems: viz. Addis Ababa, Central, Western, Southwestern, Northwestern,Eastern, Northern and Northeastern regions. In addition,the power transmission network is extended to Djibouti in the east and to the Republic of Sudan to the northwest.The HV transmission network consists of 1071.76 km of 400 kV circuits, 5895.54 km of 230 kV circuits and 4666.79 km of 132 kV circuits [1].

In Ethiopia, the power system’s vulnerability has worsened from time to time because of the implicit and explicit reasons. However, these reasons have not been studied in a concrete way which further exacerbate the situation and as a result partial and total blackouts frequently happen on the grid. Therefore, this paper presents an approach for assessing vulnerability of the EEP system by using two indices, called performance indices(PIs) which reflect the health of the system. The indices allow assessing two different symptoms of system stress such as voltage limit violations and overloads.

This process of vulnerability analysis involves studying the effect of the removal of a system component on the system, particularly the power flows and the bus voltages.This component can be a generator, transmission line,transformer, and so on. This work however focuses on outages of the transmission network. The reason is that most of the previous partial and total blackouts happened as results of the faults on the transmission lines, as reported by National Load Dispatch Center of EEP (NLDC) in [2].What follows is a qualitative and quantitative evaluation of which line outages lead to the most issues with regard to voltage limit and line load violations.

The method used to rank vulnerabilities is taken from[3], [4], [5], and [6]. Detailed explanation is given in Section II.

2 Vulnerability Assessment Methods

2.1 Performance Indices

The system performance index is a measure that can be used to evaluate the relative severity of a contingency[3]. The most common form of system performance indices gives a measure of the deviation from rated values of system variables such as line flows, bus voltages, bus power injections, etc. The indices to quantify problems related to loading and voltage limit violations are described in equations (1) and (2).

1) Active power performance index (PIρ): This index is used to measure the degree of line over loads and is given by [4].

Fig.1 Line outages ranking in terms of PIρ

where

Pl – The MW flow of line l

Pl max – The MW capacity limit of line l

Nl – The number of overloaded lines in the system

W – Real power weighting factor (in general, W = 1)

n – A positive number (n = 1, 2, 3… etc...).

The summation is carried out on overloaded lines only to avoid the masking problem reported in [3], [4], [5].

2) Voltage performance index (PIv): The voltage level performance index chosen to quantify system deficiency due to out of limit bus voltages in defined by [4].

where

Vi – The voltage magnitude corresponding to bus i

Vi SP – Specified voltage magnitude corresponding to bus i

n – A positive number and usually its value is 1 (n = 1, 2, …etc...).

NB – Number of buses in the system whose voltage magnitude is out of the specified ranges.

w – Real non-negative weighting factor (in general w =1).

Thus, this index measures the severity of the out of limit bus voltages and provides a direct means of comparing the relative severity of the different line outages on the system voltage profile.

2.2 Line Outage’s Severity Ranking

Ranking is performed for all branch outages against their rated power carrying capacity (i.e., a load of 100%),and considering voltage violations outside the range of 0.9 pu to 1.1 pu. The most severe line outages and the performance indices for the two criteria are shown in descending order in Fig.1 and 2.

Fig.1 shows the ranking of line outages in terms of the performance index calculated based on the branches overload. It can be seen that the most dangerous line outage is the disconnection of Debre Markos_Sululta 400 kV line that leads to an increase in the path of the power supply to the high load centers from Beles hydropower plant.

On the other hand, Fig.2 shows the line outages ranking based on the performance index calculated in terms of bus voltage limit violations. It is found that the most dangerous line outages from the voltage point of view are the disconnections of lines of Alamata_Combolcha 230 kV,Akaki I_Kality I 230 kV line I & II, and Sebeta I_Kality I 230 kV lines.

Fig.2 Line outages ranking in terms of PIv

3 Result Analysis and discussions

3.1 Analysis Results for the Most Severe Line Outages

The above proposed algorithm is used for the snapshot of the EEP grid at 19:00 on July 17, 2016. At this snapshot,the system contained 63 synchronous generators, 104 load buses, 134 substations, 99 two-winding transformers, 26 three-phase transformers, 3 shunt capacitor banks, 45 shunt reactors and 142 transmission lines. The detailed system data is taken from national load dispatch center (NLDC),power system planning department. At this specific time,the system was loaded to 1539.74 MW from which 100 MW was exported to the Republic of Sudan and 50 MW was exported to Djibouti. The full AC power flow is run to get the steady state results, such as bus voltages and line flows. Next, line outage analysis is done by considering a load scaling factor of 1.33, as this factor is used by the planning and operation department of NLDC for daily peak load forecasting and generation dispatch scheduling.This data is fed to DIgSILENT PowerFactory software and computer simulation is then performed for each line outage.

Simulation of the line outages begins with ascertaining a steady state solution of the system at base case, and it is obtained that there is overloading of components.Moreover, there are no voltage limit violations at the load buses.

Next, computer simulation is run for each line outage and, it is obtained that there are 28 severe line outages whose active power flow performance indices are greater than 1.0 pu and voltage level performance indices are greater than 0.02 pu. The ends of the transmission lines associated with the most severe line outages are labeled in Fig.3. It is evident that most of these line outages are predominantly along the highest voltage transmission lines(400 kV & 230 kV) connecting the regional power systems of the country to the capital, Addis Ababa. Almost all of the line outages affect the high load centers, i.e., Addis Ababa and central regions, as can be seen in Table 1 and Table 2. This knowledge implies that as far as the transmission network is concerned, the main weaknesses of the EEP grid are on the lines feeding Addis Ababa and Central regions from the rest of the regional power systems.

Fig.3 Portions of the EEP grid showing points of interest

Table 1 and Table 2 show loading and voltage limit violations for the most severe line outages of power system elements and buses respectively.

Table 1 Loding violations per case

Line outage(Name)Overloaded Component(s)Loading (in %)Contingency loading Base case Combolcha 230/132 kV Tr. 204.1 98.6 Alamata_Combolcha 230 kV line 149.7 39.8 Legetafo_Ayat GIS 132 kV line 137.6 92.9 D/Markos_Sululta 400 kV Cotobie_Ayat GIS 132 kV line 137.0 92.3 Ghedo 230/132 kV tr. 110.9 98.2 Combolcha_Legetafo 230 kV line 107.7 17.2 B/Dar 400/230 kV Tr1. 102.0 35.2 B/Dar 400/230 kV Tr2. 102 35.2 Sebeta I_Mekanisa 132 kV line 164.2 81.4 Cotobie_Weregenu TP 132 kV line 141.9 90.7 Kality I_Sebeta I 230 kV line Gefersa_Addis North 132 kV line 120.3 85.3 Kality I_Akaki I 132 kV line I & II 119.9 96.1 Combolcha 230/132 kV Tr. 113.4 98.6 Gefersa_Kality I 132 kV line 109.6 25.5

continue

Sebeta I_Mekanisa 132 kV line 164.2 81.4 Cotobie_Weregenu TP 132 kV line 141.9 90.7 Kality I_Akaki I 230 kV line I, II Gefersa_Addis North 132 kV line 120.3 85.3 Kality I_Akaki I 132 kV line I, II 119.9 96.1 Combolcha 230/132 kV Tr. 113.4 98.6 Gefersa_kality I 132 kV line 109.6 25.6

Continue

Line outage(Name)Overloaded Component(s)Loading (in %)Contingency loading Base case Cotobie_Addis East II 132 kV line 124.4 36.0 Gefersa_Addis North 132 kV line Kality I_Akaki I 132 kV line I, II 105.0 96.1 Combolcha 230/132 kV Tr. 104.1 98.6 Sebeta I_Mekanisa 132 kV line 101.4 81.4 Combolcha_Legetafo 230 kV line Combolcha 230/132 kV Tr. 128.7 98.6 Kality I_Akaki I 132 kV line I, II 100.1 96.1 Cotobie_Weregenu TP 132 kV line Kality I_Weregenu TP 132 kV line 147.2 52.6 Kality I_Akaki I 132 kV line I, II 105.6 96.1 Legetafo_Ayat GIS 132 kV line Gefersa_Addis North 132 kV line 130.4 85.3 Kality I_Akaki I 132 kV line I, II 119.1 96.1 Cotobie_Addis East II 132 kV line Gefersa_Addis North 132 kV line 129.9 85.3 Cotobie_Weregenu TP 132 kV line 104.3 90.7 Gefersa_Sebeta I 230 kV line Kality I_Akaki I 132 kV line I, II 109.7 96.1 Combolcha 230/132 kV Tr. 104.7 98.6 G/Gibe II_Sebeta II 400 kV line Sekoru 230/132 kV Tr. 106.6 72.6 Combolcha 230/132 kV Tr. 102.4 98.6 Sebeta II_Akaki I 400 kV line Combolcha 230/132 kV Tr. 102.7 98.6 Kality I_Akaki I 132 kV line I, II 101.7 96.1 G/Gibe II_Sekoru 400 kV line Metu 230/66/15 kV Tr. 101.5 95.5 Combolcha 230/132 kV Tr. 100.8 98.6

The result shows several violations as indicated in Table 1 & 2. With these violations, it is unlikely that the system would be able to operate. Voltages outside their limits could lead to widespread instability, failures, and damage of equipment while overloads could permanently damage equipment, also leading to instability [3]. Normally,procedures need to be put in place to prevent the system from reaching such an operating point and may require that part of the system get shut down in a coordinated way to save the system from damage, at the cost of interrupting supply to some customers. Such type of cutting of supply to customers results in forced blackouts, and are very common in the Ethiopian electricity sector and have been widely reported by NLDC in [2] and [8], though scientific publications are not available.

Table 2 Voltage limit violations per case

Line outage(Name) Bus name (s)Voltage (p.u)Actual Base case Alamata_Combolcha 230 kV line Akaki I 132 kV 0.90 0.92 Sululta 230 kV 0.89 0.93 Kality I 230 kV 0.89 0.93 Legetafo 230 kV 0.88 0.92 Sululta 400 kV 0.88 0.93 Gefersa 132 kV 0.87 0.91 Kality I 132 kV 0.87 0.90 Shehedi 230 kV 0.87 0.90 Sululta 132 kV 0.87 0.91 Combolcha 230 kV 0.82 0.94 Combolcha 132 kV 0.81 0.93 D/Berhan 132 kV 0.84 0.91 Akaki I_Kality I 230 kV line I, II Kality I 230 kV 0.89 0.93 Legetafo 230 kV 0.89 0.92 Shashemene 132 kV 0.89 0.92 Debrezeit TP 132 kV 0.88 0.92 Debrezeit II 132 kV 0.88 0.92 Akaki I 132 kV 0.88 0.92 Shoarobit 132 kV 0.88 0.91 Sululta 132 kV 0.87 0.91 Awasa 132 kV 0.87 0.90 Gefersa 132 kV 0.87 0.91 D/Berhan 132 kV 0.87 0.91 Kality I 132 kV 0.86 0.90 Kality I_Sebeta I 230 kV line Kality I 230 kV 0.89 0.93 Legetafo 230 kV 0.89 0.93 Shashemene 132 kV 0.89 0.92 Debrezeit TP 132 kV 0.89 0.92

Continue

?

Continue

?

3.2 Determination of Vulnerable Branch Components and Buses

Generally, results of the line outage simulation give an idea about vulnerable buses whose voltages must be maintained within limits by transmission and substation rehabilitation and upgrading (TSRUP) projects. For instance, by installing static shunt compensators at high load buses or by implementing under voltage load shedding(UVLS) protection schemes, voltage instability can be improved. Overloading of lines and/or transformers can be reduced by increasing the capacities the lines and transformers.

Network’s vulnerable buses are those buses that have out of limit voltages in case of line outages. Accordingly,the network’s most vulnerable buses are Sululta and Gefersa 132 kV buses in which 11 different line outages cause them to have out of limit bus voltages. The second most vulnerable buses are Kality I and D/ Berhan 132 kV buses where 9 different line outage cases lead them to have out of limit bus voltages. Fig.4 shows the network’s most vulnerable cases ranked based on the number of outages that caused voltage limit violations starting from the most vulnerable buses.

Fig.4 Vulnerabilities ranked based on number of outages that lead to voltage limit violations of buses

Similarly, results of line outage analysis can also give us an idea about the vulnerable equipment (lines, transformers or generators) whose capacity must be increased by TSRUP to withstand load violations as a result of line outages,and to secure operation during contingencies. Network vulnerabilities with respect to component’s overloading are the lines/transformers which always become overloaded in case of different line outages.

Based on the analysis, it is obtained that the network’s most vulnerable component is the transformer at Combolcha II 230/132 kV substation, in which 12 different line outage cases lead this transformer to overload. Table III shows the network vulnerabilities ranked based on number of line outages lead to lines/transformers overload starting with the most vulnerable component besides each component’s highest percentage load.

To sum up, as can be seen from Table 3 and Fig.4,most of the network vulnerabilities in respect of voltage limit violations and loading violations of buses and components respectively, are occurred on those buses and components either found at the high load centers or they are major power flow paths to it.

Table 3 Vulnerabilities ranked based on number of line outages lead to component overloads

Highest percentage loading Combolcha 230/132 kV Tr. 12 204.1 Kality I_Akaki I line I, II 12 119.9 Gefersa_Addis North 132 kV line 6 129.9 Cotobie_Weregenu TP 132 kV line 5 141.9 Overloaded components Number of outages lead to overload Sebeta I_Mekanisa 132 kV line 4 164.2 Gefersa_Kality I 132 kV line 3 109.6 Ghedo 230/132 kV Tr. 2 213.7 Legetafo_Ayat GIS 132 kV line 2 137.6 Cotobie Ayat GIS 132 kV line 2 137.0 Sekoru 230/132 kV Tr. 2 133.1 Alamata_Combolcha 230 kV line 1 149.7 D/Markos 400/230 kV Tr. 1 135.7 Cotobie _Addis East II 132 kV line 1 124.4 Sululta 400/230 kV Tr1 & Tr2 1 117.6 Combolcha_Legetafo 230 kV line 1 107.7 B/Dar 400/230 kV Tr1 & Tr2 1 102.0 Metu 230/66/15 kV Tr. 1 101.5

4 Conclusions

This paper mainly describes a framework for identifying the Ethiopian electric power system vulnerabilities, largely based on performance indices and supported by power flow and line outage analyses.

In the first step, ranking and identification of most severe line outages are done based on performance indices.With these identified line outages, network vulnerabilities are evaluated based on:

Numbers of outages that lead to voltage limit violations of buses.

Numbers of outages that lead to overload.

The highest percentage load of components.

Accordingly, the analysis shows that most of the network vulnerabilities in respect of bus voltage violations and element overloading are occurred on those buses and elements found at the high load centers.

This paper does not only identify the vulnerable lines,transformers and buses but also gives clear technical information related to how these vulnerabilities can be mitigated.