II Electric Power Industry in Indonesia

3. Analysis of Distribution Line Interruption

As the project scope involved a highly extensive geographical area, 1 branch was selected for each subsidiary and interruption details were analyzed. APJ Semarang of PLN Central Java Regional office, APJ South Surabaya of PLN East Java Regional office and AJ South Bali of PLN Bali Regional office were selected and the entire interruption details for each of the 3 branches were collected from the branch offices respectively and analyzed.

3-1 Semarang

Semarang branch office which is located in Central Java is the biggest branch office in that region. It possesses not only the most distribution facilities but also major public offices and important customers such as regional government offices. It is the core office in Central Java regional distribution office. Many outages happen in this area and it has a difficulty in operating distribution facilities because of this reason. We expect to improve the reliability of total Central Java regional office by analyzing the outage status of Semarang branch office which has the most interruptions and establishing the countermeasures in the final report hereafter.

3-1-1 Outage Statistics Management

It is rather difficult to implement outage statistics management on site because of unclear outage code definitions in the outage classification structure. Therefore, APJ Semarang should analyze the sources of faults in detail and establish strategies to address them in order to manage outage statistics, but, first of all, it is necessary to review and overhaul outage statistics management system and interruption classification structure itself at PLN head office level.

As for outage statistics management of APJ Semarang, DCC (Distribution Control Center) prepares fault log in reference to the outage classification structure of PLN and submits the log to the branch. Then, the branch receives the log and manages it in reference to their own separate classification codes to make it easier and simpler to classify the sources of faults so that fault operational controller can analyze faults more efficiently. However, even the classification structure at branch level do not classify faults more specifically by source, facility or area but simply breaks down faults by item, which implies room for more improvement. Outage statistics management is important for electrical utilities because their ultimate goal is to analyze outage statistics systematically and accurately, develop strategies to prevent faults with analytical data, prevent outages in advance and improve supply reliability in the end. APJ Semarang manages outage statistics adequately on their own but does not go as far as utilizing outage statistics to develop fault prevention strategies. Therefore, The section concerning recommendation for improvement approach herein will propose an interruption classification structure and statistics management techniques. It will be more sensible for PLN to customize and utilize such classification structure in line with its own requirements rather than accepting it as it is.

3-1-2 Monthly Interruption Analysis

We compared fault details records from January to December, 2006 to analyze the interruption occurrence status of APJ Semarang and found that the fault log of DCC and the fault records of APJ Semarang were well in order. DCC operators log fault descriptions and CB trip details on their applicable working day and submit the log to applicable department of relevant branch. Fault operational controller at branch also consolidates data on a monthly basis and maintain fault statistics systematically. However, as the fault log of DCC does not describe the source of fault clearly, it may raise some difficulties for fault operational controller to maintain fault statistics and analyze faults. In addition, cooperation with DOC (Distribution Operation Center) is more important than anything else for DCC to record the sources of faults in their fault log accurately and specifically. Therefore, to identify source of fault clearly, DOC staffs who actually rush to fault sites must make efforts to identify the source voluntarily and PLN must extend proactive support. In case of sustained interruption from October when rainy season starts outages increase rapidly and After April when rainy season finishes interruptions decrease gradually.

Sustained interruption happened 215 cases in November of 2006 which were the most frequent number and possessed 14% of total interruption of the year 2006. In September it happened 40 cases which were the least number and possessed 3% of total interruption of the year 2006. However in case of temporary interruption even though the number of outages in rainy season was somewhat higher than the number of outages in dry season the occurrence of temporary interruption has not increased unlike the sustained interruption. Generally temporary interruption happens much more than sustained interruption. In case of Semarang branch office it is different from ordinary case.

It is necessary to grasp and survey this reason more in detail when we write the final report. In case of temporary interruption in 2006 it occurred 31 cases in January and February which were the most frequent number and possessed 12% of total interruption. Similarly In September it happened 10 cases which were the least number and possessed 4% of total interruption of the year 2006.

[Table 2-1] Monthly interruption status in Semarang branch office(2006)

[Table 2-2] Comparison of Interruption between Semarang B/O and KEPCO(2006)

Temporary interruption happens generally more than sustained interruption in most of utilities. When fault caused by tree contact, bad weather such as high wind, debris contact happens most circuit breakers in substation are operated by fast curve because touching time between wire and tree or debris is relatively short time. Therefore usually it causes temporary interruption. As shown in the table above in case of KEPCO 8,214 temporary interruptions out of total 9,585 interruptions have happened in 2006. It took 86%. In case of sustained interruption 1,371 cases have happened and it possessed only 14%. However in case of Semarang branch office 265 temporary interruptions out of total 1,795 interruptions have happened in 2006. It only took 15%. In case of sustained interruption 1,530 cases have happened and it possessed as much as 85% on the contrary. We can assume that once the fault happen in Semarang branch office most of fault cause the sustained interruption based on this statistics. Therefore it need to be surveyed the exact reason of fault and analysed the data more thoroughly. With this continuous efforts we can reduce the fault and increase the reliability. In addition, not only protect coordination between circuit breaker in substation and protect devices in distribution network but also close and organic cooperation between distribution control center and distribution operation center should be well prepared for the fault recovery.

3-1-3 Interruption Analysis by Cause

The fault cause and fault point are mixed in the fault record of Semarang branch office.

Moreover the meaning of some items is not clear and obscure. In general when the fault is surveyed it is categorized into fault cause and fault point. After that the countermeasures for fault prevention are established based on that result. Thus the fault analysis of Semarang branch office somewhat needs to be improved in the future. For example, in case of ‘jumper broken’ and ‘wire broken’ fault point was recorded but it didn’t mention why it was broken. Load shedding, Overload and CB incoming trip, it is a little difficult to find out the reason and point of fault when we analyze the interruption data.

If we examine the total interruption data of 2006 in Semarang branch office the number of item ‘Unknown’ was 1,013 cases. It possessed 59.62% out of total 1,795 interruption cases and occupied the largest portion. The second largest item was ‘hit by tree’ which was 166 cases and occupied 9.77%. The third was ‘FCO fuse’ which was 150 cases and possessed 8.33%. One important fact is that the item ‘Unknown’ reached almost 60% out of total faults. It must be studied thoroughly in order to reduce the unknown fault and needs more efforts to find out the exact reason of fault.

3-1-4 Interruption Analysis by Substation and Feeder

The largest faults that were 223 cases and occupied 12.43% happened in GIS Kalisari substation out of total 13 substations in Semarang branch office. The faults in P. Lamper substation were 202 cases and possessed 11.26% which was the second largest portion. The third was faults in T. Lorok substation which were 177 cases and possessed 9.87%. We can find out that many faults have happened in many substations in Semarang branch office even though the number of interruption in each substation is a little different.

Total number of distribution feeder in Semarang branch office is 102 feeders. The number of feeder which haven’t experienced an outage in 2006 was 15 feeders and the other 87 feeders that were approximately 85% of total feeders have experienced the interruption at least more than once. As mentioned above the total number of yearly interruption is 1,530 cases in case of sustained interruption, and 265 cases in case of temporary interruption. Accordingly 1,795 cases occurred totally in 2006 and when it is calculated by interruption time the electricity has failed for 61,027 minutes. If it is converted into hours it is about 1,017 hours.

3-1-5 Interruption Analysis by Interruption Frequency and Duration

The result of upper 5 feeders that have experienced by interruption frequency and interruption time. In connection with interruption frequency the number of outage in Kpk.1 feeder of Krapyak substation was no less than 67 cases in one year and the duration of outage was 1,002 minutes. This means outages happened about 6 times in every month and furthermore occurred
more than once in every week on the average. Not to mention it tens of faults in a year have happened in other upper feeders such as KDO.2 feeder and PWI.8 feeder, etc. It requires that the electric company try to improve the ability of maintaining distribution facilities and preventing faults. In addition when the interruption by feeder is reviewed 62 cases of interruption in a year have happened in KDO.2 feeder of KEDUNG OMBO substation and the duration of interruption was 4,055 minutes at that time. The duration of the other feeders was also around 2,000 minutes. It was quite frequent times and long duration. We found that many PLN engineers did their best to progress the reliability. However, regardless of this situation the outage analysis and establishment of countermeasures for fault prevention are needed more constructively.

3-2 Surabaya

Surabaya branch office which is located in East Java is the biggest branch office in that region. Surabaya is the second largest city in Indonesia. Distribution facilities and distribution operation system are well equipped in Surabaya branch office that is also the core office in East Java regional distribution office like Semarang branch office in Central Java. Surabaya branch office has more distribution facilities and personnel than those of Semarang branch office. In connection with interruption analysis we have studied the interruption status of South Surabaya branch office. We expect to improve the reliability of total Surabaya area based on the analysis of outage status in South Surabaya branch office.

3-2-1 Outage Statistics Management

Unlike APJ Semarang, APJ Surabaya Selatan of PLN Dist. Eastern Java Surabaya uses interruption classification codes that PLN uses currently as they are. Therefore, it is somewhat difficult to develop fault prevention strategies as fault statistics are based on the existing code structure. As the case is the same as APJ Semarang, fault operational controller at APJ Surabaya Selatan was not aware of clear code definitions and some code definitions were considerably ambiguous. In addition, APJ Surabaya Selatan outsources fault recovery and fault statistics collection following recovery to a 3rd party contractor and receives monthly fault status and recovery data from such contractor. Therefore, considering the dependence on a 3rd party contractor for fault recovery and statistics management, it is most important to manage such contractor adequately. Such an approach may be sensible as long as a 3rd party contractor provides its service adequately. However, if it is otherwise, accuracy of statistics will be at risk.

Therefore, even if a 3rd party contractor is responsible for fault recovery and monthly fault log maintenance, the branch itself needs to receive fault data from DCC and identify/analyze the sources of faults. In addition, it also seems necessary to compare the branch’s own data with data availble from the 3rd party contractor because it is still necessary to verify the fault records from a 3rd party contractor anyway. However it is still true that APJ Surabaya Selatan could manage its 3rd party contractor better in relation to fault statistics management. APJ Surabaya Selatan receives data from its contractor and manages outage statistics but do not go as far as utilizing such data to develop fault prevention strategies. Therefore, it is inevitable to manage fault classification structure more clearly and specifically in the future and leverage it actively in developing fault prevention strategies.

3-2-2 Interruption Analysis by Month

Like Semarang branch office in Central Java outages increase gradually from October when rainy season starts and the most interruptions happened in January. After that the outages have decreased and less interruptions have occurred in dry season from April to September. If we compared sustained interruption with temporary interruption, temporary interruptions happened less than sustained interruptions unlike other electric power companies in the world.

Generally temporary interruption happens much more than sustained interruption. In case of South Surabaya branch office it is different from ordinary cases. It is necessary to grasp and survey this reason more in detail when we write the final report. In addition since most outages have happened in rainy season it is required to make an effort to reduce the faults and establish the measures for fault prevention in preparation for rainy season.

[Table 2-3] Monthly interruption status in South Surabaya branch office

[Table 2-4] Comparison of Interruption between South Surabaya B/O and KEPCO(2006)

Temporary interruption happens generally more than sustained interruption in most of utilities. When fault caused by tree contact, bad weather such as high wind, debris contact happens most circuit breakers in substation are operated by fast curve because touching time between wire and tree or debris is relatively short time. Therefore usually it causes temporary interruption. In case of South Surabaya branch office 257 temporary interruptions out of total 686 interruptions have happened in 2006. It only took 37%. In case of sustained interruption 429 cases have happened and it possessed as much as 63% on the contrary. We can assume that once the fault happen in South Surabaya branch office most of fault cause the sustained interruption based on this statistics. Therefore it need to be surveyed the exact reason of fault and analysed the data more thoroughly. With this continuous efforts we can reduce the fault and increase the reliability.

If we examine the total interruption data of 2006 in South Surabaya branch office Both sustained interruption and temporary interruption have happened most times in January. In case of temporary interruption it happened 52 cases. It possessed 20% out of total temporary interruption cases. In case of sustained interruption it happened 72 cases. It possessed 17% out of total sustained interruptions. In October only 7 temporary interruptions which was the lowest in 2006 happened and 20 sustained interruptions happened in December.

3-2-3 Interruption Analysis by Cause

If we examine the total interruption data by interruption code in 2006 specifically, the code ’41c-MV CB or fuse because of rain, lighting’ was 154 cases. It possessed 41% out of total interruption cases which occupied the largest portion. The second largest item was ’41d-MV wire because of other cause’ which was 48 cases and occupied 12%. The third was ’41e-Relay is working without clear cause’ which was 36 cases and possessed 9%. Like the other branch offices the largest portion of interruption cause in South Surabaya branch office is the item ‘Unknown’. Thus it must be studied thoroughly in order to reduce the unknown fault and needs more efforts to find out the exact reason of fault.

3-2-4 Interruption Analysis by Substation and Feeder

In connection with interruption frequency the number of outage in ANGKASA PURA feeder of RUNGKUT substation was 19 cases that wat the most frequent cases in one year and the total duration of that outage was 678 minutes. This means outages happened about twice in every month and we can assume that there would be some complaints from customers. It will influence not only the profit for the electric company but also bad image of the company. It requires that the electric company try to improve the ability of maintaining distribution facilities and preventing faults. Furthermore, the outage analysis and establishment of countermeasures for fault prevention are needed more actively.

The largest faults that were 149 cases and occupied 35% happened in RUNGKUT substation out of total substations in South Surabaya branch office. The faults in SUKOLILO substation were 75 cases and possessed 17.5% which was the second largest portion. The third was faults in DARMO GRAND substation which were 75 cases and possessed 17.5%. We can find out that 66% of total faults has happened in those 3 substations of South Surabaya branch office.

Thus when the maintenance program or inspection plan will be made, these substation should be proceeded first.

3-3 Bali

In case of Bali distribution regional office South Bali branch office has been selected for the outage analysis. South Bali branch office which includes major attraction places for tourists such as Denpasar, Kuta and Sanur is the core branch office in that region. There are totally 13 substations in Bali and South Bali branch office has 7 substations out of them. It has also 94 distribution feeders. Because Bali distribution regional office has facilities much fewer than the other distribution regional office, it is a little advantageous to maintain the distribution facilities compared to the other regional offices. Accordingly we have found that South Bali branch office has managed well systematically about interruption record and interruption statistics management. However, if we take the underground facilities in Bali, interruptions in South Bali branch office happened rather many times. So more accurate and thorough outage analysis is proceeded in order to work out outage prevention program and countermeasures for reliability improvement in the future.

3-3-1 Outage Statistics Management

Unlike APJ Semarang, APJ Surabaya Selatan of PLN Dist. Eastern Java Surabaya uses interruption classification codes that PLN uses currently as they are. Therefore, it is somewhat difficult to develop fault prevention strategies as fault statistics are based on the existing code structure. As the case is the same as APJ Semarang, fault operational controller at APJ Surabaya Selatan was not aware of clear code definitions and some code definitions were considerably ambiguous. In addition, APJ Surabaya Selatan outsources fault recovery and fault statistics collection following recovery to a 3rd party contractor and receives monthly fault status and recovery data from such contractor. Therefore, considering the dependence on a 3rd party contractor for fault recovery and statistics management, it is most important to manage such contractor adequately. Such an approach may be sensible as long as a 3rd party contractor provides its service adequately. However, if it is otherwise, accuracy of statistics will be at risk.

Therefore, even if a 3rd party contractor is responsible for fault recovery and monthly fault log maintenance, the branch itself needs to receive fault data from DCC and identify/analyze the sources of faults. In addition, it also seems necessary to compare the branch’s own data with data availble from the 3rd party contractor because it is still necessary to verify the fault records from a 3rd party contractor anyway. However it is still true that APJ Surabaya Selatan could manage its 3rd party contractor better in relation to fault statistics management. APJ Surabaya Selatan receives data from its contractor and manages outage statistics but do not go as far as utilizing such data to develop fault prevention strategies. Therefore, it is inevitable to manage fault classification structure more clearly and specifically in the future and leverage it actively in developing fault prevention strategies.

3-3-2 Interruption Analysis by Year and Month

In case of PLN Bali regional office we have analyzed the interruption data of South Bali branch office in 2006. When the yearly interruption data of South Bali branch office from 2003 to 2006 have been reviewed the results. It is desirable that generally the outages have been decreasing. Specifically the interruption has decreased from 1,441 cases in 2003 to 1,032 cases in 2006. Even though it is very difficult to reduce the faults greatly within short period of time for 2 or 3 years, the survey showed that south Bali branch office made a remarkable performance so far. We analysed that it is the result of making an active effort to reduce the faults not only from the regional office but also the branch office.

When monthly interruption data of south Bali branch office in 2006 have been reviewed it is similar to that of other branch offices. From October when rainy season starts outages increase slightly and After April when rainy season finishes interruptions decrease gradually. Sustained interruption happened 134 cases in January of 2006 which were the most frequent number and in October it happened 46 cases which were the least number.

[Table 2-5] Yearly and Monthly interruption status of South Bali branch office

3-3-3 Interruption Analysis by Cause

The interruptions relater to ’41d-MV wire because of other cause’ were 35 cases in 2006. It possessed 17% out of total interruption cases which occupied the largest portion. The next largest items were ’41a-MV wire because of tree or kite’ which was 23 cases and occupied 11.4%. The third was ’49-Others / MV overhead wire’ which was 10 cases and possessed 10%.

Like the other branch offices the largest portion of interruption cause in South Bali branch office is the item ‘Other cause’. Thus it must be studied thoroughly in order to reduce the unknown fault and needs more efforts to find out the exact reason of fault.

The interruption code ’41d, 41e, 49′ have taken large portion of cause in temporary interruption. However, When we examine this result, the meaning of these codes is ‘unknown’.

In addition, the cause of interruption ‘unknown’ also frequently appeared in the lower items.

Thus it must be studied thoroughly in order to reduce the unknown fault and needs more efforts to find out the exact reason of fault like sustained interruption.

3-3-4 Interruption Analysis by Substation and Feeder

There are 94 distribution feeders in south Bali branch office and there are 63 distribution feeders that it never happened interruption in 2006. 67% of distribution feeder out of total distribution feeders has experienced interruption and in case of temporary interruption it has reached 85%. From this fact we can assume that interruption happened broadly in south Bali branch office. If we consider that the scale of distribution facilities in south Bali branch office is less than the other branch office, many interruptions have happened. Thus the preparation and countermeasures are also needed to reduce the faults.

In case of sustained interruption 201 cases have happened in 2006 and total duration of interruption time was 7,596 minutes, that is 126.6 hours. If we examine the total outage statistics by substation in South Bali branch office, the largest faults that were 55 cases happened in Kapal substation out of total substations in South Bali branch office. The faults in Pesanggaran substation were 51 cases which were the second largest portion. The third was faults in Nusa Dua and Sanur substation which were 33 cases. In case of temporary interruption it happened 831 cases. The total duration of interruption time was 1,097 minutes, that is 18.3 hours. The largest faults that were 268 cases happened in Kapal substation out of total substations in South Bali branch office like sustained interruption. The faults in Pesanggaran substation were 185 cases which were the second largest portion. The third was faults in Sanur substation which were 142 cases. Based on this data most interruptions have happened in high density of load area like downtown and major attraction places for tourists.

Most of the in interruption in South Bali branch office happened in Kapal substation, Pesanggaran substation and Sanur substation.

If we examine the most frequent feeder of interruption in south Bali branch office, Penebel feeder in Kapal substation has happened 10 cases in 2006 and it means almost every month has 1 case of outage. 12 feeders have had 6 interruptions in a year. In case of temporary interruption 38 cases has happened in Express Seminyak feeder of Padangsambian substation and we can assume that more than 3 temporary interruptions happen in a month. More than 20 temporary interruptions in a year have happened in 12 feeders including Suwung feeder of Pesanggaran substation.

3-4 Consolidation of Fault Analysis

3-4-1 Semarang

1. Analyzing faults thoroughly & developing fault prevention strategies
   APJ Semarang manages fault statistics well in general. However, They do not classify and analyze faults by cause or facility and simply maintains the number of faults per each fault, which results in insufficiency of raw data supporting development of fault prevention strategy later on and difficulties with establishing effective fault prevention strategy. For example, the fact that unknown causes account for more than half of the faults seems to imply insufficient effort to identify the cause of fault. Therefore, it is necessary to define causes of faults more specifically and encourage maintenance crews to identify causes of faults actively on site.
2. Patrolling & inspecting distribution lines efficiently
   Line patrol and inspection is very important in preventing faults in distribution lines as faults that can happen in the future can be prevented in advance by patrolling and inspecting distribution lines thoroughly. Therefore, distribution branches must patrol and inspect lines thoroughly as scheduled. It is still true that APJ Semarang does not patrol and inspect lines effectively. It is our judgment that the current operational staffs alone cannot handle the workload practically, considering that significant number of faults occur in not only high-voltage lines but also in low-voltage lines as well.

   As the operation center of APJ Semarang is already busy with fault recovery alone, line patrol and inspection plan must be established on a separate track. Fault recovery is important, but, faults must be prevented in advance by rigorous line patrol and inspection. As fault prevention also leads to reduction of fault recovery workload, it is like catching two birds with a stone.

3. Recovering faults promptly
   As is the case with KEPCO, APJ Semarang outsources most of its fault recovery operation to a 3rd party contractor. However, since spare parts and materials for emergency maintenance are not available unlike KEPCO, it is likely that fault recovery will take significantly long. For example, transformer failures occur quite frequently due to heavy air-conditioning load under tropical weather conditions. Movable transformer truck is used in transformer recovery, but, since APJ Semarang has only 1 movable transformer truck, difficulties are anticipated if two or more transformers fail concurrently.
4. Calculating reliability indices accurately
   PLN maintains SAIFI and SAIDI, primary reliability indices widely used around the world.
   Accordingly, not only APJ Semarang of PLN Dist. Central Java but also APJ Surabaya Selatan of PLN Dist. Eastern Java and AJ Bali Selatan of PLN Dist. Bali maintain SAIFI and SAIDI.

   APJ Semarang does not maintain any objectives concerning the number of faults separately, but, makes a lot of efforts to improve SAIFI and SAIDI since they fall into the scope of management target. However, there is still some room for further improvement in terms of their efforts to calculate such indices accurately. SAIFI and SAIDI must be calculated based on not only high-voltage interruptions but also low-voltage interruptions and scheduled outages as well to ensure accuracy. In case of scheduled outage, accurate number of customers connected to interrupted sections must be considered in calculation. However, APJ Semarang neither reflects scheduled outage data on reliability index calculation nor identifies the exact number of customers connected to interrupted sections during scheduled outage, which results in insufficient facility history data and calls for more improvement in the future.

3-4-2 Surabaya

1. Analyzing faults thoroughly & developing fault prevention strategies
   APJ Surabaya Selatan also manages fault statistics well in general. However, as is the case with APJ Semarang, they do not classify and analyze faults by cause or facility and simply maintains the number of faults per each fault, which results in insufficiency of raw data supporting development of fault prevention strategy later on and difficulties with establishing effective fault prevention strategy.

   Furthermore, APJ Surabaya Selatan needs to manage fault statistics more rigorously as they outsource fault reporting and analysis to a 3rd party contractor. If competent staff does not pay enough attention or manage data adequately, errors in data available from the contractor may have significant impacts on the operation of PLN facilities. Fault analysis should not just stop as such, but, it is most important to leverage fault analysis to develop fault prevention strategies to reduce faults. Faults must be reduced incrementally by analyzing each fault item thoroughly and developing approaches to improve the quality of device/material if material quality is an issue or suggesting fault prevention guideline for rainy season if faults occur frequently in rainy seasons. In addition, fault analysis and prevention activities conducted in a year must be reviewed in the following year to review the effectiveness of such activities comprehensively every year. A feedback system described above is expected to be effective in reducing trials and errors consistently and contributing to the development of optimum fault analysis and prevention system.

2. Patrolling & inspecting distribution lines efficiently
   Industrial area covers most part of Surabaya given its local characteristics. As the air pollution is heavier than other regions because of the emission from industrial plants, power system facilities will deteriorate faster than other areas. Therefore, APJ Surabaya Selatan must patrol and inspect lines more frequently and efficiently than other areas as scheduled. However, it still seems true that APJ Surabaya Selatan does not patrol and inspect lines effectively. It is our judgment that the current operational staffs alone cannot handle the workload practically, considering that significant number of faults occur in not only high-voltage lines but also in low-voltage lines as in the case of APJ Semarang.
3. Recovering faults promptly
   As is the case with KEPCO, APJ Surabaya Selatan outsources most of its fault recovery operation to a 3rd party contractor. However, since spare parts and materials for emergency maintenance are not available unlike KEPCO, it is likely that fault recovery will take significantly long.

   In addition, if a fault occurs in a certain line, load impacted by the faults are not adequately transferred to other lines in case of APJ Surabaya Selatan as extra line capacity is not sufficient.
   DCC operators also do not feel comfortable switching loads freely even in case of a line with 2 interconnection points as they are concerned if interconnected lines may be tripped simultaneously, which indicates that it is necessary to review distribution line interconnection and division configuration comprehensively in order to raise the efficiency of distribution line operation.

4. Calculating reliability indices accurately
   APJ Surabaya Selatan also maintain SAIFI and SAIDI. APJ Surabaya Selatan does not maintain any objectives concerning the number of faults separately, but, makes a lot of efforts to improve SAIFI and SAIDI since they fall into the scope of management target. However, there is still some room for further improvement in terms of their efforts to calculate such indices accurately. SAIFI and SAIDI must be calculated based on not only high-voltage interruptions but also low-voltage interruptions and scheduled outages as well to ensure accuracy. In case of scheduled outage, accurate number of customers connected to interrupted sections must be considered in calculation. However, APJ Surabaya Selatan neither reflects scheduled outage data on reliability index calculation nor identifies the exact number of customers connected to interrupted sections during scheduled outage, which results in insufficient facility history data and calls for more improvement in the future.

3-4-3 Denpasar

1. Analyzing faults thoroughly & developing fault prevention strategies
   AJ Bali manages fault statistics well in general. However, just like other branches, AJ Bali does not classify and analyze faults by cause or facility and simply maintains the number of faults per each fault, which results in insufficiency of raw data supporting development of fault prevention strategy later on and difficulties with establishing effective fault prevention strategy.

   However, AJ Bali Selatan attempts not to use “Unknown Cause” as a fault code to improve fault recovery practice since significant number of faults were attributed to “Unknown Cause”, which indicates commitment to identifying cause of fault on site one way or another. Removing “Unknown Cause” from the fault code list will provide a good momentum to change the slack mindset of some DOC staffs who may neglect the responsibility for identifying the cause of fault. With or without success, such attempts are deemed to be very important and worth benchmarking by other branches.

2. Patrolling & inspecting distribution lines efficiently
   As AJ Bali Selatan’s service area is small relative to other branches, it is easier to patrol and inspect distribution lines. However, as the natives in Bali have strong emotional attachment to kites, it is always possible to see a kite in the sky. Electrical utilities must be cautious about kites as kites can trigger faults of power line and AJ Bali Selatan must tighten line patrol and inspection more rigorously than other branches. But in spite of such environmental constraints, AJ Bali Selatan suffers considerable fewer faults than other branches. The reason seems to be because AJ Bali not only patrols and inspects lines but also manage facilities more efficiently in general than other branches.
3. Recovering faults promptly
   As is the case with KEPCO, AJ Bali Selatan outsources most of its fault recovery operation to a 3rd party contractor. In addition, AJ Bali Selatan also operates a separate mobile maintenance team for fault recovery. The mobile maintenance team arrives at fault site on motor cycles fast even through traffic jam during commuting hours. Such fault recovery system merits extra compliment from customer service perspective and may minimize customer damage and public inconvenience and contribute more to increasing the revenue of energy sales by reducing outage hours.

   However, unlike KEPCO, AJ Bali Selatan does not have separate inventory of emergency maintenance materials, therefore, fault recovery is likely to suffer delay. Emergency maintenance materials, if any, are mostly miscellaneous materials and primary components and materials such as transformer or power line wire are not available to AJ Bali Selatan.

   In addition, if a fault occurs in a certain line, load impacted by the faults are not adequately transferred to other lines in case of AJ Bali Selatan as extra line capacity is not sufficient, which calls for a comprehensive review of line interconnection and division configuration to improve efficiency of distribution line operation.

4. Calculating reliability indices accurately
   PLN maintains SAIFI and SAIDI, primary reliability indices widely used around the world.
   Accordingly, AJ Bali Selatan maintains SAIFI and SAIDI. APJ Semarang does not maintain any objectives concerning the number of faults separately, but, makes a lot of efforts to improve SAIFI and SAIDI since they fall into the scope of management target. However, there is still some room for further improvement in terms of their efforts to calculate such indices accurately.

   SAIFI and SAIDI must be calculated based on not only high-voltage interruptions but also low-voltage interruptions and scheduled outages as well to ensure accuracy. In case of scheduled outage, accurate number of customers connected to interrupted sections must be considered in calculation. However, AJ Bali neither reflects scheduled outage data on reliability index calculation nor identifies the exact number of customers connected to interrupted sections during scheduled outage, which results in insufficient facility history data and calls for more improvement in the future. AJ Bali now calculates reliability indices manually now, but, needs to introduce and utilize a reliability indices calculation program in the future.