Risk Assessment Framework for Power Circuit Breakers Based on Condition, Replacement, and Criticality Indices

Abstract

This paper develops a comprehensive framework for the risk assessment of 115 kV power circuit breakers (PCBs) by evaluating their condition, replacement needs, and criticality to the electrical network. The primary objective is to create a risk assessment tool that enhances maintenance practices and improves operational efficiency. The framework begins with a condition assessment, quantified through the use of a health index, derived from historical diagnostic test results and routine checks. The next step involves a replacement assessment, using a replacement index that considers factors such as age, rating adequacy, and technological obsolescence to determine the necessity of replacement. Finally, a criticality assessment is performed using a criticality index, which evaluates the PCB’s role in the network by factoring in location, load importance, failure severity, and the consequences of failure on network operations. By integrating these indices, the framework offers a holistic view of the associated risks. The methodology is applied to assess the risk of 149 sample PCBs across 30 substations in Thailand, with relevant data collected for each unit. The resulting risk assessments support proactive maintenance, minimize downtime, optimize the allocation of limited resources, and enhance the overall efficiency, reliability, and safety of the electrical network.

1. Introduction

A power circuit breaker (PCB) is an essential component in electrical networks as it is designed to make, carry, and break the electrical current under both normal and abnormal conditions to maintain system stability and network reliability. In high voltage systems, particularly 115 kV and above, PCBs often employ SF₆ gas as an arc-quenching medium and insulation in their interrupting units [1,2]. Many PCBs used in Thailand’s electrical system have a live tank design that enables the replacement of major components. Over time, these PCBs experience aging and degradation due to various factors, such as mechanical and electrical stress, seal hardness, inadequate maintenance, and the operational environment. This degradation can reduce the operational capability of PCBs and increase the likelihood of failure, which poses a significant challenge to the reliability and safety of electrical networks, particularly in preventing unplanned outages [3,4,5].

Several studies have explored these degradation mechanisms and proposed models to predict and mitigate their effects. For example, several studies have analyzed how different stress conditions accelerate equipment deterioration and have introduced predictive maintenance models to address these challenges [6]. These models typically focus on the probability of failure (PoF) as a key element in the risk assessment, derived from historical data, degradation modeling, and physical inspections [7,8,9,10]. Various probabilistic models, such as Weibull analysis, Bayesian networks, and Monte Carlo simulations, are commonly used to forecast the failure rates of aging equipment [11,12,13]. PCBs with higher failure rates normally require earlier replacement than those with lower failure rates. Accurate determination of the optimal replacement time supports the development of effective inventory and maintenance strategies. Moreover, financial losses associated with unforeseen cascading failures can be mitigated [14].

Diagnostic techniques for evaluating PCBs are essential to ensuring the reliability and safety of power systems. Standards such as IEC 62271-100 [15] and IEEE C37.09 [16] provide guidelines for testing the electrical, mechanical, and dielectric performance of PCBs. These tests include dielectric breakdown, contact resistance, timing, and interruption performance evaluations. Advanced methods such as vibration monitoring and thermal imaging are increasingly used to detect mechanical degradation and overheating, which are common signs of wear and failure [17]. Data from condition monitoring tools, including electrical testing, thermal imaging, and vibration analysis, have been used to derive health indices for PCBs [18,19,20]. These indices provide a quantitative measure of the PCB’s operational status, offering insight into degradation trends and supporting maintenance and replacement decisions. Recent advancements in condition-based monitoring have improved real-time data collection, enabling continuous health assessments. Composite health scores, incorporating parameters such as contact resistance, timing, and insulation resistance, provide a more comprehensive evaluation of the PCB’s condition and degradation. Multi-criteria decision-making (MCDM) methods, such as the Analytical Hierarchy Process (AHP) [21], help systematically evaluate various factors to support effective maintenance decisions [22,23,24,25].

To investigate the condition, risk, and criticality associated with PCBs, several methodologies have been reviewed [26,27,28,29,30]. The AHP expert-based method [26] provides structured multi-criteria prioritization by leveraging expert judgment to identify critical PCBs. However, it relies heavily on subjective input, which can lead to inconsistency and limits its scalability across large networks. In [27], a data-driven hybrid model employing fuzzy C-means clustering and Support Vector Machine (SVM) ranking enhances objectivity and effectively manages incomplete data. However, it requires clean, well-labeled training datasets and offers limited interpretability for practical field applications. In [28], influence diagrams with interval probabilities were used to support probabilistic reasoning and reduce expert bias through the use of credal sets. However, this method demands probabilistic expertise due to its mathematical complexity, making it impractical for routine utility operations. The adaptive backpropagation neural network model [29] enables dynamic, nonlinear condition assessment, but functions as a black box system, requiring large datasets and expert tuning, limiting transparency. Real-time life-cycle assessment using control circuit monitoring [30] provides accurate, condition-based maintenance planning, but is costly and depends on widespread online monitoring infrastructure. Additionally, recent works [31,32,33] emphasize predictive maintenance and fault modeling through analytics and MCDM frameworks, improving decision making, but often requiring advanced tools or hybrid integration that may not be universally feasible.

To address the limitations of previous studies, this paper proposes a risk-based maintenance assessment that integrates condition, replacement need, and criticality evaluations. Using actual diagnostic and operational data, along with weighting and scoring techniques and the AHP, the methodology is used to compute a Health Index (HI), Renovation Index (RI), and the probability of failure (PoF) for PCBs. As illustrated in Figure 1, the proposed approach enables simple, transparent, replicable, and data-driven decisions by incorporating failure probability, asset replacement urgency, and criticality. This supports optimized maintenance scheduling, the prioritization of critical assets, and an extended PCB lifespan, ultimately improving the performance and reliability of power systems.

2. Health Index and Probability of Failure Determination

The Health Index (HI) represents the actual condition of a PCB, derived from testing and diagnostic results. This index is sensitive to varying environmental and operational factors, such as high salinity in coastal or marine environments, high humidity in tropical regions, and extreme heat and dryness in arid areas, which can accelerate deterioration and directly affect the HI.

The testing and inspection methods used by a utility in Thailand are summarized in

Table 1. Criteria and scoring for health assessment of 115 kV power circuit breakers.

Criteria	Good (0)	Moderate (3)	Poor (5)	Remark
Insulation Resistance (GΩ)	≥1		<1	The worst score is used as a representative for further evaluation
Contact Timing Test (%Deviation)	≤10		>10
Contact Resistance (μΩ) (Model Dependent)	≤200, 40		>200, 40
SF6 Pressure Inspection	Normal	Alarm #1	Alarm #2
SF6 Dew Point (Temperature, °C)	<−10		>−10
SF6 Percentage (Amount of SF6, %)	≥80		<80
SF6 Purity (%)	≥99		<99
Amount of SO2 (ppm)	≤10		>10
Visual Inspection	Normal		Abnormal
Gas Leakage Rate (times/year)	≤1		≥3

. The criteria, along with weighting and scoring techniques, are used to assess each PCB’s condition. Each criterion is classified into three categories: good (0), moderate (3), and poor (5). The category ranges are defined using equipment manuals, international standards, and expert judgment to accurately reflect the PCB’s condition. The criteria are generally defined as follows:

• Insulation resistance: values ≥ 1 GΩ indicate good insulation, lower values increase leakage and breakdown risks;
• Contact timing: a deviation ≤10% from the commissioning test results indicates good performance, a higher deviation suggests mechanical issues and overheating risks;
• Contact resistance: values ≤200 μΩ indicate good conductivity, higher values suggest wear, overheating, or failure risk;
• SF₆ pressure: normal pressure ensures proper functioning, low pressure signals leakage or malfunction;
• SF₆ dew point: a value below −10 °C means dry gas is present, higher values indicate humidity, reducing insulation reliability;
• SF₆ percentage: a value >80% ensures good insulation, lower levels suggest depletion and failure risk;
• SF₆ purity: >99% purity ensures dielectric strength, lower purity increases failure risk;
• SO₂ level (ppm): a value <10 ppm is normal, higher levels suggest gas decomposition from internal arcing and damage;
• Visual inspection: detects physical defects like cracks or rust, abnormalities require repair or replacement;
• Gas leakage rate: ≤1 leak/year indicates proper sealing, >3 leaks/year suggests poor sealing and environmental concerns.

To calculate the Health Index (%HI), which reflects a PCB’s condition, all the evaluation criteria are assigned weights using the AHP, a decision-making method based on pairwise comparisons that is used to rate their relative importance on a scale from 1 (equal importance) to 9 (extreme importance). These comparisons form matrices from which eigenvalues are calculated to derive the weights. A consistency check ensures the logic of the comparisons. The weighted scores of all the criteria are then combined to compute the %HI, ranking the PCB’s condition based on measurable performance. Thus, the weighting factors, evaluation criteria, and test/inspection scores enable the conversion of quantitative data into a %HI value.

The worst score approach for evaluating the %HI was chosen to emphasize the importance of preventive and risk-based maintenance in ensuring electrical system reliability. In regard to these systems, a single point of failure can cause significant consequences, such as equipment damage or service disruption. Prioritizing the component in the poorest condition prevents degradation from being masked by healthier components and highlights critical vulnerabilities. Unlike averaging or fuzzy methods that may dilute risk indicators, the worst score strategy ensures that urgent issues are clearly identified, supporting timely and effective maintenance. Once scoring and weighting are obtained, the %HI is calculated using Equation (1).

$$\%HI={\displaystyle \frac{{\displaystyle \sum _{i=1}^{10}\left(S_i\times W_i\right)}}{{\displaystyle \sum _{i=1}^{10}\left(S_{\max ,i}\times W_i\right)}}}\times 100,$$

(1)

where

• S_i = Scoring value of each criterion for the health assessment;
• S_max,i = Maximum score of each criterion for the health assessment;
• W_i = Weighted value of each criterion for the health assessment;
• i = Number of criteria for the health assessment.

The probability of failure (%PoF) measures the likelihood that a system or component will fail within a specific time or under certain conditions. It is expressed as a percentage, with 0% indicating no risk and 100% indicating certain failure. The %PoF complements the %HI, which reflects the equipment’s condition, and is often calculated using Equation (2).

$$\%PoF=1-\%HI$$

(2)

where %HI is a measure of the equipment’s operational health.

This formulation of the %PoF is based on the assumption of linear degradation, a widely accepted and practical approach used in asset management and reliability studies. It provides a transparent and empirically supported method for condition-based risk assessments.

A high %PoF indicates that an asset is nearing the end of its life and is at an increased risk of failure. The %PoF plays a critical role in condition-based maintenance (CBM) and Reliability-Centered Maintenance (RCM), supporting the prioritization of maintenance and replacement activities to minimize downtime and enhance system performance.

3. Replacement Index Determination

The invisible aging of PCBs is assessed using the criteria listed in

Table 2. Key criteria for replacement assessment of 115 kV power circuit breakers.

Criteria	Low (0)	Moderate (3)	High (5)	Wi
Overall Age	<10	11–24	≥25	30.60
Age of Interrupter	<22	23–24	≥25
Age of Mechanism after Overhaul	<10	11–24	≥25
Number of Fault Current Interruptions	<4	5–8	≥9
Ratio of Load Current to the Rating of the Equipment	<0.6	0.6–0.7	≥0.8	21.00
Ratio of Short-Circuit Current to the Rating of the Equipment	<0.6	0.6–0.7	≥0.8
Already Replaced by Advanced Technology	No	-	Yes	8.20
No Longer Manufactured	No	-	Yes
Spare Parts Availability	Easy to find	Difficult to find, possible to modify	Unable to modify	20.00
Personnel Expertise Level	Good	Moderate	Poor
OEM Support/After-sale Service Quality	Good	Moderate	Poor
Number of Units Remaining in Use	≥5	-	<5
Operator Level of Satisfaction (Failure Rate)	Satisfied	-	Not satisfied
Major/Minor Failure, Cost and Effort to Repair	Minor defect	-	Major defect	20.0

. These criteria, with defined scoring ranges and weights, calculate the replacement index (%RI), which reflects the breaker’s remaining performance and replacement need. The scores are categorized into three levels, low (0), moderate (3), and high (5), based on organizational standards and maintenance expertise. The highest severity score among all the criteria is used for the final replacement decision to avoid overlooking critical issues due to averaging. The key criteria for assessing replacement needs and determining the %RI are defined as follows:

• Overall age: ≥25 years indicates high failure risk and urgent replacement, <10 years reflects low risk, 11–24 years requires regular inspection and testing;
• Interrupter age: ≥25 years signals degraded interruption performance and high failure risk, requiring replacement;
• Mechanism age after overhaul: ≥25 years suggests end-of-life status and the need for major refurbishment;
• Fault current interruptions: more than nine interruptions indicates severe contact wear, requiring detailed interrupter inspection;
• Load current to rating ratio: a ratio >0.8 means that operation is near capacity, suggesting the need for relocation or a higher rated replacement;
• Short-circuit current to interrupting capacity ratio: a ratio >0.8 indicates high electrical stress, requiring immediate replacement to avoid failure;
• Technology obsolescence: incompatible or outdated technology justifies replacement or a system upgrade;
• Manufacturing status: discontinued models with unavailable parts should be replaced with current alternatives;
• Spare parts availability: limited availability suggests end of life and supports replacement or the need for major refurbishment;
• Personnel expertise: a lack of trained personnel for maintenance increases risks and supports the use of more maintainable equipment;
• OEM support: the absence of OEM technical or parts support often necessitates replacement or refurbishment;
• Similar units remaining: fewer than five similar units in operation complicates support and part sourcing, justifying replacement;
• Operator satisfaction: frequent failures or negative feedback reflects poor reliability and supports upgrades or replacement;
• Repair cost and effort: high repair costs or maintenance effort suggests that replacement may be more cost effective.

After determining the evaluation criteria and their scoring values, the weights for the 14 criteria shown in Table 2 were assigned using the AHP, as applied in regard to the HI determination. Subsequently, the %RI of the PCB can be calculated using Equation (3).

$$\%RI={\displaystyle \frac{{\displaystyle \sum _{l=1}^{14}\left(S_l\times W_l\right)}}{{\displaystyle \sum _{l=1}^{14}\left(S_{\max ,l}\times W_l\right)}}}\times 100,$$

(3)

where

• S_l = Scoring of each criterion for the replacement assessment;
• S_max,l = Maximum score of each criterion for the replacement assessment;
• W_l = Weighted value of each criterion for the replacement assessment;
• l = Number of criteria for the replacement assessment.

4. Criticality Index Determination

The criticality index (CI) of a PCB reflects both the likelihood of failure and the severity of its impact. The criteria, categorized into two groups, as detailed in

Table 3. Key criteria for criticality assessment of 115 kV power circuit Breakers.

Criteria	Low	Moderate	High	Wi
1. Possibility of Equipment Failure
1.1 Short-Circuit Current (kA)	<10	10–20	>20	2.559
1.2 Pollution level (Bird Droppings, Sea Spray)	No		Yes	2.559
2. Severity/Consequence
2.1 Function of Substation	Supply Load		Switching/Terminal	3.089
2.2 Bus Arrangement	Main and Trans. H bus	Double Main and Trans.	Breaker and a Half	3.089
2.3 Number of Circuits	<4	4–7	>7	1.828
2.4 Number of Power Transformers	1	2	0, 3	1.828
2.5 Available Area for Future Expansion	Greater Area	Limited Area	Unable to Expand	0.898
2.6 Safety (Distance from Community)	>3 km	1–3 km	<1 km	0.898
2.7 Public Image	No		Yes	0.898
2.8 Substation Location (Load Importance)	Rural, District	Province	City, Industrial Estate, Tourist Area	2.694
2.9 Loading Percentage	<60	60–80	>80	33.82
2.10 Function of Bay	Tie Bay	Transmission Line	Transformer/Line from Power Plant	29.13
2.11 Redundancy/Planned Outage	Yes/Long Duration	Sometimes/Short Duration	No	15.71

, along with their corresponding weights determined through brainstorming and the AHP, are used to calculate the percentage criticality index (%CI). This index assesses the importance of a substation bay by evaluating the probability and consequences of failure, aiding the prioritization of maintenance, upgrades, and replacements. By quantifying equipment criticality, the %CI enables utilities to prioritize the most essential assets, thereby improving system reliability and operational efficiency. The criteria are defined as follows:

• Possibility of Equipment Failure
  • Short-circuit current: the maximum current during a fault. Currents below 10 kA pose low risk, while those above 20 kA indicate high electrical stress and greater failure risk.
  • Pollution level: Environmental contaminants like bird droppings or sea spray can degrade insulation and cause flashovers. Higher pollution levels increase equipment criticality.

• Severity/Consequence of Equipment Failure
  • Function of substation: Substations serving load supply are less critical, while switching or terminal substations are more critical due to their role in power flow control.
  • Bus arrangement: Redundancy in bus configurations reduces failure risk. H-bus or main and transfer schemes offer low redundancy, while the breaker-and-a-half substation provides the highest reliability and criticality.
  • Number of circuits: indicates network importance. Fewer than four circuits suggests low criticality, more than seven indicates high criticality.
  • Number of power transformers: More than three transformers signifies higher capacity and greater system importance.
  • Available area for future expansion: Sufficient space reduces criticality by allowing future upgrades, limited or no space increases it.
  • Safety (distance from community): Substations near urban or industrial areas pose higher safety risks and are more critical.
  • Public image: Substations in visible, populated, or tourist areas are more critical due to public and environmental concerns.
  • Substation location (load importance): rural locations are less critical; urban, tourist, or industrial locations are highly critical due to their role in the operation of key infrastructure.
  • Loading percentage: above 80% indicates high operational stress and criticality, below 60% suggests low stress.
  • Function of bay: tie bays are least critical; transformer and transmission line bays are highly critical for power distribution.
  • Redundancy/planned outage: high redundancy and tolerance for outages reduce criticality; systems with low tolerance are highly critical.

After establishing the evaluation criteria, the weights and scores for the 13 criteria, as shown in Table 3, were assigned using the AHP, as applied in regard to the HI determination. The %CI of the PCB can then be calculated using Equation (4).

$$\%CI={\displaystyle \frac{{\displaystyle \sum _{k=1}^{13}\left(S_k\times W_k\right)}}{{\displaystyle \sum _{k=1}^{13}\left(S_{\max ,k}\times W_k\right)}}}\times 100$$

(4)

where

• S_k = Scoring value of each criterion for the criticality assessment;
• S_max,k = Maximum score of each criterion for the criticality assessment;
• W_k = Weighted value of each criterion for the criticality assessment;
• k = Number of criteria for the criticality assessment.

5. Results and Discussion

The risk assessment of the PCBs was conducted by calculating the %PoF, %RI, and %CI, based on actual test results and operational data collected by a state utility in Thailand. Due to data-sharing restrictions, the raw data are not included in this manuscript. However, the results and conclusions presented in this study are grounded on operational data and reflect a practical, experience-based assessment and analysis. The results are presented in Figure 2 and Figure 3. A two-dimensional (2D) plot of the %PoF versus the %RI illustrates the relationship between visible aging (%PoF) and invisible aging (%RI), which reflects the replacement needs. By incorporating the %CI into this 2D plot, a three-dimensional (3D) visualization is created, allowing for a more comprehensive assessment of the PCB’s condition. This 3D plot assists in identifying high-risk equipment and prioritizing maintenance or replacement actions based on both the condition of the PCBs and their criticality within the system.

5.1. Results of Probability of Failure, Replacement Index, and Criticality Index

This section presents the practical data collected from 149 PCB units installed across 30 substations in Thailand, as shown in

Table 4. Practical data of 115 kV PCBs (%PoF, %CI, %RI, No. of Units).

(%PoF, %CI, %RI)	No. of Units	(%PoF, %CI, %RI)	No. of Units	(%PoF, %CI, %RI)	No. of Units	(%PoF, %CI, %RI)	No. of Units
(6, 35.51, 12)	1	(15.18, 43.95, 30.36)	2	(27.66, 40.50, 55.32)	1	(69.28, 23.86, 38.56)	3
(6, 40.70, 12)	2	(15.18, 46.43, 30.36)	1	(27.66, 52.15, 55.32)	2	(69.28, 26.73, 38.56)	2
(6, 42.74, 12)	1	(15.18, 47.93, 30.36)	3	(27.66, 52.15, 55.32)	1	(69.28, 38.38, 38.56)	1
(6, 43.95, 12)	2	(15.18, 54.16, 30.36)	2	(56, 20.73, 12)	1	(69.28, 41.79, 38.56)	1
(6, 44.66, 12)	2	(15.18, 54.39, 30.36)	8	(56, 32.38, 12)	1	(69.28, 44.67, 38.56)	4
(6, 52.35, 12)	1	(15.18, 55.60, 30.36)	2	(56, 39.37, 12)	1	(69.28, 47.16, 38.56)	1
(6, 59.60, 12)	6	(15.18, 59.59, 30.36)	9	(56, 44.03, 12)	1	(69.28, 53.44, 38.56)	3
(9.18, 23.86, 18.36)	1	(15.54, 48.16, 31.08)	2	(56, 46.79, 12)	1	(69.28, 56.32, 38.56)	2
(9.18, 35.51, 18.36)	1	(19.28, 21.09, 38.56)	1	(56, 47.70, 12)	1	(71.3, 44.31, 42.6)	1
(9.18, 47.16, 18.36)	2	(19.28, 22.523, 38.56)	2	(56, 50.95, 12)	1	(71.3, 56.32, 42.6)	1
(9.18, 49.26, 18.36)	6	(19.28, 27.90, 38.56)	1	(56, 61.27, 12)	12	(75.64, 49.44, 51.28)	1
(9.18, 60.91, 18.36)	6	(19.28, 44.39, 38.56)	1	(57.66, 28.84, 55.32)	1	(75.64, 55.72, 51.28)	1
(15.18, 23.12,30.36)	1	(19.28, 45.83, 38.56)	3	(65.18, 23.48, 30.36)	1	(75.9, 53.37, 51.8)	1
(15.18, 30.86, 30.36)	1	(19.28, 50.678, 38.56)	1	(65.18, 35.14, 30.36)	2	(79.88, 31.47, 59.76)	1
(15.18, 32.29, 30.36)	1	(19.28, 51.20, 38.56)	2	(65.18, 35.51, 30.36)	1	(79.88, 49.41, 59.76	3
(15.18, 34.78, 30.36)	2	(25.64, 48.16, 51.28)	6	(65.18, 46.79, 30.36)	1	(79.88, 61.06, 59.76)	6
(15.18, 42.51, 30.36)	2	(25.64, 59.81, 51.28)	3	(65.3, 44.31, 30.6)	1	total no. of PCBs	149

. The dataset includes the calculated values of the %PoF, %RI, and %CI for each unit, providing insight into the current condition, operational reliability, and criticality of these PCBs within the electrical system. Table 4 shows that several PCBs share common values for the %PoF, %RI, and %CI, particularly in the lower ranges (e.g., %PoF = 6, %RI = 12, with 15 units) and higher ranges (e.g., %PoF = 9.18, %RI = 18.36, with 16 units). This suggests that these PCBs are likely of the same type, model, or manufacturer, or were produced in the same batch and operate under similar environmental and operational conditions. For example, PCBs with a 6% PoF and a 12% RI likely originate from the same product line, exhibiting similar behavior across multiple units under comparable use cases, climates, and loading conditions. However, variations in the %CI are observed due to differences in substation location and the operational importance of each unit within the network.

5.2. The 2D Plotting of %PoF vs. %RI

The %PoF and %RI are key indicators used to assess the condition of 115 kV power circuit breakers (PCBs). The 2D plot of the %PoF versus the %RI in Figure 2 demonstrates a clear linear relationship between these variables. The %PoF reflects visible aging, derived from physical inspections and measurable data that represent the breaker’s current physical condition. As the PCB experiences aging, wear, or erosion, the %HI declines, resulting in a higher %PoF and indicating the need for urgent repair.

In contrast, the %RI captures invisible aging by considering the operating conditions, usage patterns, rating adequacy, technology obsolescence, spare part availability, and user satisfaction. Although not immediately visible, these factors significantly impact the PCB’s long-term performance and reliability. As these conditions worsen, the %RI increases, signaling the need for replacement, even in the absence of obvious physical deterioration.

Low values in terms of both indices indicate that the PCB is in good condition, with high reliability and minimal risk of failure. In this state, only routine maintenance and periodic visual inspections are typically required. Over time, as aging and operational stress accumulate, the %PoF gradually increases, and the %RI rises correspondingly, indicating growing failure risk. As deterioration advances, more frequent inspections, closer monitoring, and potentially planned refurbishment become necessary to ensure continued reliability.

When the %RI becomes high, the PCB is approaching the end of its useful life, and the %PoF increases sharply. At this stage, the breaker becomes a high-risk asset, and failure is more likely if replacement is delayed. Postponing replacement can lead to unplanned outages, safety hazards, and increased repair costs. The %PoF versus %RI plot enables operators to determine optimal replacement timing, balancing the cost of early replacement against the risks of failure. By closely monitoring these two indices, utility operators can make informed decisions to optimize maintenance planning and asset replacement schedules.

5.3. The 3D Plotting by Adding the %CI to 2D Plotting of %PoF Versus %RI

Adding the %CI to the 2D plot of the %PoF versus %RI transforms the plot into a 3D plot, as shown in Figure 3, providing a comprehensive view of the breaker’s likelihood of failure, reliability during operation, and its importance to the electrical network. The %CI reflects the breaker’s significance within the power system by considering factors such as its role in maintaining system stability, the impact of its failure on the network, and its operational relevance. This additional dimension enables system operators to assess not only the condition and failure risk of PCBs, but also the potential consequences of their failure based on their criticality to the grid. Incorporating the %CI helps utilities prioritize the PCBs that pose both a high failure risk and high criticality. For example, a breaker with a high %PoF, %RI, and %CI would be a top candidate for replacement, due to the severe impact its failure could cause. Conversely, breakers with high failure risk but lower criticality may be assigned a lower priority. The 3D analysis facilitates a holistic decision-making approach, optimizing maintenance and replacement strategies to minimize operational risks and costs.

5.4. Discussion

The analysis of different %PoF, %RI, and %CI values has been organized into key focus areas including a failure probability assessment, operational reliability, criticality evaluation, unit distribution, system risk management, and preventive maintenance prioritization, as follows.

Regarding the failure probability assessment, the %PoF values in the dataset vary significantly. The highest values, such as 79.88%, 75.90%, and 75.64%, indicate that a substantial number of units are at high risk of failure. Units with a %PoF exceeding 70% should be prioritized for urgent inspection and corrective maintenance to mitigate potential outages. In contrast, units with lower %PoF values, such as 6%, present minimal failure risk and typically require only routine inspection and preventive maintenance.

From the reliability assessment perspective, %RI values in the dataset range from 12% to 59.76%. Lower %RI values, such as 12%, indicate higher operational reliability, suggesting that long-term replacement planning is sufficient. In contrast, higher %RI values, for e.g., 59.76%, reflect lower reliability and signal a greater likelihood of short-term failure. To ensure system reliability, attention should be focused on units with higher %RI values, as these units exhibit lower reliability during operation.

In respect to the criticality assessment, the %CI reflects the importance or impact of each unit in the electrical system. Units with higher %CI values, such as 61.06 and 59.81, are more critical and have a greater impact on system performance. These units should be prioritized first for maintenance or replacement, as their failure has significant consequences. Conversely, units with lower %CI values, for e.g., 23.86 and 20.73, are less critical, meaning their failure would have less consequences or a lower risk of supply interruptions. This distinction is essential for effective resource allocation and risk mitigation planning.

In regard to the unit distribution assessment, the data reveal a notable pattern. Certain %CI values are associated with larger clusters of units, for example, six units with %CI = 61.06 and a high %PoF of 79.88, indicating a concentration of critical, high-risk breakers. In contrast, many other %CI values correspond to only one or two units, suggesting a wide variability in criticality across the system. This distribution implies that while most PCBs have diverse roles within the network, specific groups pose concentrated risks. Identifying and monitoring these larger clusters is essential for targeted risk mitigation and maintenance planning.

In terms of system risk management, ten PCB units exhibit the highest %PoF values, such as 79.88%, as shown in Table 1, indicating a substantial risk of failure at these specific locations. To address this risk effectively, priority should be given to the six units with the highest %CI value of 61.06%, as they present a high failure probability, with significant system impact. Conversely, units with both a low %PoF and low %CI pose minimal risk and can be deprioritized. This risk-based approach enables more efficient allocation of resources, focusing on the most critical and vulnerable assets to reduce the overall risk to system reliability.

Focusing on preventive maintenance, the results support a strategic maintenance plan that prioritizes units with both high %PoF and %CI values. This includes units with a %PoF above 70% (e.g., 79.88%, 75.90%) and a %CI above 50% (e.g., 61.06%), which pose the greatest risk and should be targeted for immediate inspection, servicing, or replacement. In contrast, units with lower %PoF and %CI values below 30% can be maintained through routine inspections and scheduled maintenance, as their likelihood of failure and impact on system performance are minimal. This risk-based prioritization approach enables the efficient allocation of resources, while enhancing overall system reliability. The method has been validated through practical implementation within a utility environment, demonstrating its applicability and effectiveness during practical operating conditions.

In addition, the method is designed to leverage existing diagnostic and operational data routinely collected by utilities, thereby minimizing the need for additional data gathering efforts. Moreover, ongoing advancements in data acquisition technologies and asset monitoring systems are expected to further ease data collection burdens in the near future, enhancing the feasibility of comprehensive risk-based maintenance strategies.

6. Conclusions

The analysis of the %PoF, %RI, and %CI for a fleet of 149 PCBs in Thailand’s 115 kV distribution network provides valuable insights into system performance and risk assessment. These results are derived from actual test data and operational records collected by a state utility in Thailand, reflecting a practical, experience-based evaluation.

The 2D analysis of the %PoF and %RI offers an initial understanding of the probability of failure and operational reliability of PCBs. This information helps identify PCB units that are more likely to fail and are less reliable during operation. This analysis serves as a foundation for identifying vulnerable units, but it does not account for the criticality of PCBs in regard to the overall power system. To bridge this gap, incorporating the %CI into a 3D analysis, alongside the %PoF and %RI, provides a more comprehensive view of PCB usage risk. This analysis not only considers failure probability and operational reliability, but also accounts for the criticality of each PCB unit within the distribution network.

By incorporating the %CI, utilities can prioritize PCBs that are both likely to fail and that are highly critical to the system’s operation, enabling immediate action where needed. In this way, limited human and financial resources can be allocated more effectively. PCB units with high %PoF and %CI values should be prioritized for urgent inspection and servicing, while those with lower values can undergo regular preventive maintenance. This strategic approach enables optimal resource allocation, reduced system risks, improved reliability, and more effective cost management in maintaining PCBs within the electrical system.

The proposed method offers a balanced, practical, and scalable solution for utilities seeking to prioritize circuit breaker maintenance and replacement, using operational data and well-structured indices. It bridges the gap between high-level strategic planning and field-level decision making, without depending on expensive technology or complex probabilistic models.