Advancing Sustainability in the Power Distribution Industry: An Integrated Framework Analysis

Abstract

This study examines the efficiency of Pakistan’s power distribution industry through an index that is experiencing financial and technical losses resulting in poor service quality, blackouts, and high tariffs. The index reveals a moderate decline from 2007–2015 and a decline to poor levels by 2022, with some improvement noted in reducing distribution losses and increasing recoveries. However, certain DISCOs have seen a decline in indicators such as reliability, quality service, safety, and recoveries, requiring continuous improvement. The study proposes a variety of measures to enhance the distribution sector’s performance, such as underground distribution, energy auditing, compliance with safety standards set by OSHA, addressing fuel scarcity to reduce load shedding, implementing smart metering and prepaid metering, and developing information technology infrastructure to interact with consumers.

1. Introduction

Electricity is a crucial component in determining the economic development of both emerging and transitional economies [1]. However, the distribution of electricity from the transmission system to consumers, the final stage of energy delivery, is often perceived as the weakest link in the power sector due to its high susceptibility to disruptions and technical and economic losses [2]. As the demand for electricity and generation capacity continues to grow, the distribution power sector faces new challenges that threaten the financial sustainability of utilities and hinder economic progress [3]. Operating and managing the electricity industry efficiently and affordably to satisfy the ever-increasing demand for electricity has been a challenge for power utilities in emerging economies [1]. Recoverable amounts from consumers, distribution losses, delays in tariff notifications, subsidies, and the reliability and quality of services are some of the main challenges faced by these economies [4].

Inefficient recoveries and losses are the primary causes of the cash flow gap in the power sector [3]. For example, in 2019, the distribution sector in India owed INR 418.81 billion to generation companies and other creditors, with an irrecoverable bill amount of INR 961.9 billion [5]. Similarly, in Pakistan, “Circular Debt” refers to the shortfall amount of PKR 1145 billion in recoveries and PKR 600 billion in distribution losses in 2020 [6]. In the past, Pakistan’s distribution sector has undergone various schemes and initiatives to enhance the operations and financial health of distribution utilities [7,8,9]. As a result, nine distribution utilities known as “DISCOs” and a regulator called the “National Electric Power Regulatory Authority (NEPRA)” were established in 1997 [10]. NEPRA was responsible for managing the power sector and ensuring its smooth operation [11]. Furthermore, NEPRA was tasked with providing reliable and high-quality power to consumers at an affordable price by conducting a “Performance Evaluation Report (PER)” [12].

Pakistan’s distribution sector is facing financial stress due to debt accumulation, with DISCOs struggling to collect revenue and achieve high recovery rates, leading to a circular debt of PKR 2.8 trillion in 2021, accounting for 52% of total debt [13,14,15]. The country’s power system has been plagued by inefficiencies such as overloaded transformers, feeders, and poor pole and line conditions, resulting in frequent power outages and an unreliable power supply [4,16]. Despite the increasing demand for electrical connections due to population growth, DISCOs have faced bureaucratic culture, corruption, and technological impediments in providing connections within specified time frames [3,17]. Load shedding has been utilized to regulate the supply-demand gap and improve the DISCOs’ performance [12], with DISCOs that have lower rates of electricity theft, lower line losses, and high recovery rates experiencing shorter blackouts [11,18].

In Pakistan, it appears that implementing regulatory standards to enforce corporate discipline and responsibility has not produced significant outcomes, despite efforts to do so [19]. While the National Electric Power Regulatory Authority (NEPRA) has conducted empirical studies, known as Performance Evaluation Reports (PERs), to assess the performance of DISCOs, they have only used data from a single year [20,21]. This approach overlooks the need for both technical and administrative evaluations to analyze the performance of distribution utilities. As a result, these evaluations appear inadequate for capturing long-term efficiency trends.

Moreover, there is a paucity of literature that comprehensively examines the performance of Pakistan’s energy distribution utilities through empirical research. Few studies have been discovered that discuss the issues related to DISCOs in Pakistan. For instance, a study [22] utilized stochastic frontier analysis to empirically investigate the economic effectiveness of energy distribution utilities in Pakistan. The authors emphasized the significance of including service quality in the regulation of energy distribution companies to avoid inefficient resource allocation. However, the weakness of the benchmarking model and variable specification cannot be ignored as they have a significant impact on the efficiency scores, and different benchmarking techniques cannot be expected to yield similar results [23]. Therefore, it is crucial to select the appropriate methodology for a benchmarking analysis based on the research question and data availability.

Several studies have been conducted to provide policy recommendations for Pakistan’s power sector. One such study, referenced as [3], specifically examined the country’s power industry and analyzed the causes of the power crisis, its financial implications, and the role of distribution utilities in contributing to the crisis. However, a significant limitation of this study was that it did not adequately address policy recommendations related to the distribution sector, which is a crucial aspect in overcoming Pakistan’s power crisis.

Other studies, such as [4], evaluated the electricity industry by utilizing multiple technical and administrative sectoral variables. The study revealed that there was a lack of consistency in the performance of DISCOs (Distribution Companies) in terms of distribution sector variables, with some showing satisfactory performance and others performing poorly. However, the study does not provide any concrete recommendations to address this issue, except for suggesting that DISCOs may introduce technological innovations such as smart grids and net metering for the development of a sustainable power sector.

Similarly, in a related study [24], the smart grid domain was discussed, specifically within a Distributed Energy Resources (DER) ecosystem, and a model was developed for end-to-end proofing. The study also provided a detailed analysis and evaluation of the offline training phase of the RL agent to highlight its key operational features and benefits over traditional RBAC models. However, one of the biggest disadvantages of an RL agent is that it may struggle to generalize well to unseen or slightly modified scenarios [25]. This limitation can pose challenges when deploying RL agents in real-world applications, as they may fail to adapt effectively to new situations without additional training or fine-tuning.

While the study [26] provides valuable insights into the technical efficiency and productivity of electric distribution utilities in Pakistan following major reforms in the power sector, it is important to note that the distance function approach used to estimate efficiency may have limitations compared to other methods. Furthermore, although the study identifies factors that can enhance productivity, it does not offer specific recommendations on how to implement these changes in practice.

To provide effective policy recommendations, it is essential to have a comprehensive understanding of the distribution sector, taking both short-term and long-term developments into account [20,21]. However, the evaluations conducted by NEPRA are based on a single year of data, which may be insufficient for capturing the current state of the sector. In addition, the lack of empirical techniques limits the available methods for objectively measuring the efficiency, effectiveness, and productivity of the sector. This makes it difficult to identify areas that require improvement and implement necessary reforms [22].

Therefore, investing in data-driven methods of analysis and evaluation is crucial to enhancing the efficiency, effectiveness, and productivity of the distribution sector and promoting sustainable economic growth. This will enable policymakers to make informed decisions and identify areas of improvement, leading to the implementation of necessary reforms to improve the sector’s performance.

The contribution of this study lies in its proposal of a comprehensive framework to evaluate the performance of Pakistan’s distribution sector, thus addressing the existing gap in the literature. The framework offers valuable guidance on suitable methods for measuring and strategizing the performance of DISCOs while leveraging NEPRA’s sectoral indicators and dimensions. By implementing this framework, significant improvements can be achieved, enhancing NEPRA’s institutional capacity and strengthening the regulatory capabilities of other developing economies. Overall, this study’s contribution lies in providing a practical and effective approach to evaluating and enhancing the performance of DISCOs, contributing to the advancement of the power distribution sector and regulatory practices in similar contexts.

To comprehensively evaluate the performance of DISCOs, the proposed methodology will be implemented in stages. The first step involves breaking down performance into constituent groups, consolidating relevant data, and creating sub-indices based on these groups. The resulting performance index will rate DISCO’s performance on a scale of 1 to 10 from 2001 to 2022. To determine the weight estimations for each sectoral indicator, Principal Component Analysis (PCA) will be used as it outperformed 12 other dimension reduction techniques [27]. The indicators will be extracted in different groups using multiple iterations of PCA. Five distribution industry experts will provide weights to each sectoral indicator, and the iteration outcomes will be synthesized based on criteria to produce the indices.

This method will increase the robustness of the index results, which will be cross-validated using equal weights. This methodology will bridge the gap in the existing literature by offering guidance on the most suitable approaches for quantifying and strategizing the performance of DISCOs using NEPRA’s sectoral indicators and dimensions. Effective implementation of this methodology will substantially improve NEPRA’s institutional capacity and enhance the capacity of other regulatory bodies in developing economies.

It is worth noting that the proposed framework offers a comprehensive and structured approach to evaluate the performance of Pakistan’s distribution sector, which is its main advantage. The framework includes several features, such as performance enhancement plans for each sector (electricity generation, distribution, and transmission), government and organizational support plans that are essential for execution, and technological innovations like smart metering, net metering, distributed generation, and prepaid billing.

The proposed framework in this study is unique in that it incorporates NEPRA’s sectoral indicators and dimensions and offers guidance on appropriate methods for quantifying and strategizing the DISCOs’ performance, which has not been previously attempted in the literature. By employing PCA and expert consultation, this study was able to extract the most significant sectoral indicators and assign appropriate weights to them, addressing the lack of a systematic method for selecting indicators and their weights [28]. This innovative approach improves the robustness of the results and provides a more precise evaluation of the distribution sector’s performance.

Additionally, this study’s novelty lies in applying this framework to Pakistan’s distribution sector, enabling the identification of the primary performance issues that can provide valuable insights for policymakers and regulators in the country. The approach is tailored to the unique characteristics of the Pakistani distribution sector, and the results can inform the development of policies and strategies that address the specific challenges in the sector. However, the index development procedure can be applied to a wide range of variable sets and energy sectors in any country; the indicators presented in this study are most relevant to countries such as India and Bangladesh that face similar challenges as Pakistan. If the indicators are to be used in other contexts, they may need to be customized to obtain accurate results.

The rest of the paper’s outline is as follows. The methodology for the index is discussed in the following section, followed by the results in Section 3. Section 4 contains discussion and policy recommendations, whereas Section 5 presents the conclusion.

2. Materials and Methods

As per the Performance Standards Distribution Rules (PDSR-2005), DISCOs are obligated to demonstrate adherence to sector-specific benchmarks, including metrics such as distribution network losses, revenue recovery, interruption frequency and duration indices, load-shedding duration, faults per kilometer, time frame for new connections, complaints per day, and safety standards [29]. NEPRA evaluates the DISCOs’ performance data against previous years’ performance, with a focus on these indicators.

The framework for evaluating the DISCOs’ compliance with these indicators is presented in Figure 1, with the definitions of each indicator and its potential impact on the index summarized in

Table 1. Metrics and their underlying definitions and impacts. Source: [29].

Sno	Indicator	Definition	Impact
1	Distribution losses (%)	There are two types of power system losses: technical and non-technical. Non-technical losses are generated by external activities on the power system, whereas technical losses occur naturally.	Negative: Lower value required to improve the performance
2	Revenue recovery (%)	The difference between the amount billed and the amount collected from the final consumer.	Positive: Higher value required to improve the performance
3	System average interruption frequency index (SAIFI)	The average number of outages experienced by a consumer in a given year.	Negative: Lower value required to improve the performance
4	System average interruption duration index (SAIDI)	The number of outages (in minutes) that a consumer experienced in a given year.	Negative: Lower value required to improve the performance
5	Duration of load-shedding (Hours per day)	The duration (Hours) of unavailability of electricity supply to the end user.	Negative: Lower value required to improve the performance
6	Number of faults per kilometer	Assesses the performance of distribution firms based on the number of faults that occur in a one-kilometer length of line.	Negative: Lower value required to improve the performance
7	Pending new connections (%)	It is the proportion of customers that did not receive new connections within the time limit specified.	Negative: Lower value required to improve the performance
8	Complaints per day	It is the number of complaints related to service interruption or voltage fluctuation received by the distribution companies.	Negative: Lower value required to improve the performance
9	Standards (fatal incidents per year).	It is the accumulated number of employees and public deaths caused due to accidents.	Negative: Lower value required to improve the performance

. The impact of each indicator is categorized as positive or negative, with a positive indicator leading to an improvement in the index if the indicator’s value increases, and vice versa for negative indicators. Notably, there is only one positive indicator and eight negative indicators (Table 1).

Table 2. Indicators data with basic statistics. Source: [15,29,30].

Dimension	Distribution Losses	Recoveries	Reliability				Quality		Safety
Indicators	Transmission and Distribution Losses (%)	Recovery (%)	System Average Interruption Frequency Index (SAIFI)	System Average Interruption Duration Index (SAIDI) (mins)	Load Shedding (hrs)	No. of Fault per km	Time Frame for New Connection (%)	Consumer Service Complaints (per Day)	Safety (Average Deaths per Year)
2001	36.87	85.9	267.2	9396	8.7	5.3	26.3	932	23
2002	36.11	85.9	259.4	9288	8.3	5.1	25.2	1088	23
2003	35.35	85.9	251.5	9180	8.0	5.0	24.2	1253	22
2004	34.59	85.9	243.7	9072	7.7	4.9	23.1	1314	22
2005	33.83	85.8	235.9	8965	7.3	4.8	22.0	1565	21
2006	33.07	85.8	228.1	8857	7.0	4.6	20.9	1815	21
2007	32.31	85.8	220.3	8749	6.6	4.5	19.8	2066	21
2008	31.55	85.8	212.4	8641	6.3	4.4	18.8	2117	20
2009	30.79	85.8	204.6	8534	6.0	4.2	17.7	2167	20
2010	30.03	85.7	196.8	8426	5.6	4.1	16.6	2355	20
2011	30.28	86.5	223	7713	4.6	4.1	15.7	2734	22
2012	28.93	85.2	192.6	8935	5.3	3.7	18.2	3750	18
2013	27.63	82.2	195.6	8898	5.4	4.3	13.6	4689	13
2014	26.34	86.8	129.3	8405	5.2	4.4	11.6	3726	19
2015	25.57	84.3	116.7	7094	3.1	2.3	8.5	6114	23
2016	24.47	87.4	110.6	7022	2.7	3.1	7.1	9644	18
2017	23.88	92	138.3	7016	3.2	2.1	9.2	16,240	14
2018	24.71	85.5	142.3	7582	3.4	2.9	5.7	26,943	16
2019	23.84	81	146.7	7997	2.4	3.5	7.6	12,109	18
2020	22.87	85.5	142.6	7669	2.5	3.3	9.5	11,428	16
2021	21.67	85.5	110.8	7240	1.9	2.6	4.7	11,022	16
2022	20.91	85.5	102.9	7133	1.6	2.5	3.7	12,348	15
Mean	28.9	85.7	185	8265	5.2	3.8	14.9	6247	19.1

provides the statistical details underlying the indicators’ data.

Index Development

To develop the index, the indicators were standardized since they have different units. The Z-score was employed for normalization, and then the indicators were grouped. However, before extracting the principal components using the Varimax rotation technique, it was essential to assess the suitability of the data for principal component analysis (PCA). For this purpose, Kaiser-Meyer-Olkin (KMO) and Bartlett’s Test of Sphericity were conducted, as per the guidelines [31]. The KMO score was calculated as 0.686, which exceeded the required value of 0.5. Additionally, the significance level was found to be 0.000, which was less than the required value of 0.005. These results confirmed the suitability of the data for PCA extraction, and the extraction was performed using the Varimax rotation technique through multiple iterations.

The first iteration of the PCA extraction involved setting the number of groups to the default value, resulting in two groups, as illustrated in

Table 3. Indicator Extractions via Varimax Rotation.

Iteration 1 (Default)	Groups
	1	2
Pending connections	0.989
TD losses	0.988
Load shedding hours	0.980
SAIFI	0.956
Faults per kilometer	0.936
SAIDI	0.908
Complaints per day	−0.858
Deaths per year	0.755
Recoveries		0.989
Iteration 2	Groups
	1	2	3
SAIDI	0.948
Load shedding hours	0.914
SAIFI	0.903
Faults per kilometer	0.897
Pending connections	0.892
TD losses	0.857
Complaints per day	−0.644
Deaths per year		0.912
Recoveries			0.997
Iteration 3	Groups
	1	2	3	4
SAIDI	0.929
Load shedding hours	0.910
SAIFI	0.910
Pending connections	0.881
Faults per kilometer	0.876
TD losses	0.862
Deaths per year		0.895
Recoveries			0.998
Complaints per day				−0.740
Iteration 4	Groups
	1	2	3	4	5
SAIDI	0.929
Load shedding hours	0.908
SAIFI	0.905
Pending connections	0.878
Faults per kilometer	0.876
TD losses	0.859
Deaths per year		0.899
Recoveries			0.999
Complaints per day				−0.749	-

. In subsequent iterations, the number of groups was fixed at 3, 4, and 5 for iterations 2, 3, and 4, respectively. It is noteworthy that when the number of groups was set to 5, the same four groups as in iteration 3 were obtained, indicating that additional iterations were not necessary. The rotation sum of square loadings was computed for each PCA iteration and presented in

Table 4. PCA-based sum of squared loadings.

	Groups Extracted	Number of Indicators	Rotation Sums of Squared Loadings
			Total	% of Variance	Cumulative %	Group Weights
Iteration 1	1	8	6.834	75.933	75.933	W1 = 75.933/88.197 = 0.860
	2	1	1.104	12.264	88.197	W2 = 12.264/88.197 = 0.139
Iteration 2	1	7	5.431	60.342	60.342	W1 = 60.342/94.685 = 0.637
	2	1	2.011	22.339	82.682	W2 = 22.339/94.685 = 0.235
	3	1	1.080	12.003	94.685	W3 = 12.003/94.685 = 0.126
Iteration 3	1	6	5.260	58.448	58.448	W1 = 58.448/97.756 = 0.597
	2	1	1.533	17.037	75.485	W2 = 17.037/97.756 = 0.174
	3	1	1.079	11.993	87.478	W3 = 11.993/97.756 = 0.122
	4	1	0.925	10.278	97.756	W1 = 10.278/97.756 = 0.105
Iteration 4	1	6	5.231	58.125	58.125	W1 = 58.125/97.645 = 0.595
	2	1	1.532	17.026	75.151	W2 = 17.026/97.645 = 0.174
	3	1	1.076	11.960	87.110	W3 = 11.960/97.645 = 0.122
	4	1	0.948	10.535	97.645	W4 = 10.535/97.645 = 0.107

.

To determine the weight of each group and indicator, the loadings were utilized based on the number of indicators in each group. For instance, in the first iteration, group 1 had eight indicators; hence, each indicator in this group had a weight of 0.860/8 = 0.107 or 11%. On the other hand, since group 2 had only one indicator, it had a weight of 0.139/1 = 0.139 or 14%. Therefore, the weight of every indicator was estimated for all groups extracted in different iterations using IBM SPSS version 21.

To ensure the robustness of the weight assignment process, the weights estimated through PCA were then cross-validated with the assigned weights by experts to clarify that the statistical technique was robust. The criteria were established based on two principles. First, the average weight assigned by experts was compared to the estimated weight for each indicator. If the weight assigned by experts closely matched any iteration, that iteration was deemed “suitable.” Second, the iteration with the highest number of “suitable” counts was selected for index development. In addition, priorities to assign weight were also taken into account. For example, experts agreed that to improve the efficiency of the sector, collection or recoveries should be of the highest priority, and the rest of the indicators can be considered to be of similar significance.

Out of nine iterations, iteration 1 had weights that were close to the expert’s weights assigned to each indicator in six instances (

Table 5. Weight assigned by experts and iteration suitability.

Indicators	E1	E2	E3	E4	E5	Average	Iteration 1	Iteration 2	Iteration 3	Iteration 4	Remarks
TD losses	10%	20%	15%	10%	20%	15%	11%	9%	10%	10%	Iteration 1 is suitable
Recoveries	20%	20%	15%	15%	10%	16%	14%	13%	13%	13%	Iteration 1 is suitable
SAIFI	10%	10%	15%	10%	10%	11%	11%	9%	10%	10%	Iteration 1 is suitable
SAIDI	10%	5%	5%	10%	10%	8%	11%	9%	10%	10%	Iteration 2 is suitable
Load shedding	20%	20%	15%	15%	10%	16%	11%	9%	10%	10%	Iteration 1 is suitable
Faults per kilometer	5%	10%	5%	10%	10%	8%	11%	9%	10%	10%	Iteration 2 is suitable
Pending connection	5%	10%	5%	10%	5%	7%	11%	9%	10%	10%	Iteration 2 is suitable
Complaints per day	10%	10%	10%	15%	10%	11%	11%	9%	11%	11%	Iterations 1, 3, and 4 are suitable
Deaths per year	15%	15%	10%	10%	15%	13%	11%	24%	17%	17%	Iteration 1 is suitable

). Therefore, group extraction based on iteration 1 was utilized for index development. This shows that the weight assignment process was carefully and systematically done to ensure the robustness and validity of the resulting index.

The cross-validation of assigned weights with estimated weights through PCA provided strong evidence that the PCA method was a valid and robust approach for estimating weights. The high level of agreement between the assigned weights and estimated weights indicates that the weight assignment process was transparent and defensible, with input from both experts and statistical analysis.

Overall, the use of both expert opinion and statistical analysis through PCA provided a rigorous and reliable approach for assigning weights to each indicator and developing the distribution sector index. This approach enhances the validity and reliability of the index and its potential use in various applications.

The index construction was carried out in multiple steps using different equations, as shown in

Table 6. Stepwise procedure for index development.

Steps	Equation
Step 1: Z score technique was used to normalize the data using the equation.	ƶ = $\frac{x-{\mu }_{\mathrm{ij}}}{\sigma }$
Step 2: For negative indicators, the inverse was estimated whereas the positive indicator was considered directly for Step 3.	${\mathrm{X}}_{\mathrm{ij}}=\frac{1}{Y_{\mathrm{ij}}}$
Step 3: The scaling was carried out for positive and negative indicators using equation. Resultantly, indicator values were converted between 1–10.	Φ = $\frac{10\times {Ind}_{\mathrm{ij}}}{{Max}_{\mathrm{i}}}$
Step 4: The value of each indicator from Step 3 is squared for positive and negative indicators using equation, denoted as ${\alpha }_{\mathrm{ij}}^2$.	${\alpha }_{\mathrm{ij}}^2={\mathsf{\Phi }}_{\mathrm{ij}}\times {\mathsf{\Phi }}_{\mathrm{ij}}$
Step 5: In this step, values of α2 from Step 4 were to be divided by number of indicators in each dimension (denoted as ո).	${\beta }_{\mathrm{ij}}=\frac{{\alpha }_{\mathrm{ij}}^2}{n}$
Step 6: Weight values from Step 1 were multiplied with indicators value from Step 5.	Ω $={\beta }_{\mathrm{ij}}\times {\mathrm{W}}_{\mathrm{ij}}$
Step 7: The final index was estimated using the square root of the sum values of indicators from Step 6 across each year using equation.	${\mathrm{P}}_{\mathrm{ij}}=\sqrt{\sum {\mathsf{\Omega }}_{\mathrm{ij}}}$

. There were two group indices estimated: “Loss-Reliability-Quality-Safety-Index” and “Recovery-Index.” Following the group indices, the final index, termed the “Performance index,” was computed.

Table 7. Group indices and performance index.

Year	√ΣGᵢ₁	√ΣGᵢ₂	Gᵢ₁ × W1 (Loss-Reliability-Quality-Safety-Index)	Gᵢ₂ × W2 (Recovery-Index)	Performance Index
2001	2.02	0.30	1.74	0.04	1.78
2002	2.10	0.30	1.81	0.04	1.85
2003	2.28	0.30	1.96	0.04	2.01
2004	2.37	0.30	2.03	0.04	2.08
2005	2.76	0.14	2.37	0.02	2.39
2006	2.93	0.14	2.52	0.02	2.54
2007	3.13	0.14	2.70	0.02	2.72
2008	4.68	0.14	4.02	0.02	4.04
2009	5.43	0.14	4.67	0.02	4.69
2010	7.08	0.02	6.09	0.00	6.09
2011	5.92	1.25	5.09	0.17	5.27
2012	8.01	0.82	6.89	0.11	7.01
2013	5.91	5.59	5.08	0.78	5.86
2014	23.30	1.73	4.95	0.24	5.19
2015	10.65	2.25	4.56	0.31	4.87
2016	4.79	2.68	4.12	0.37	4.49
2017	1.63	10.00	1.40	1.39	2.79
2018	1.71	0.34	1.47	0.05	1.52
2019	4.35	7.50	3.74	1.04	4.78
2020	2.89	0.34	2.49	0.05	2.54
2021	1.54	0.34	1.32	0.05	1.37
2022	1.3	0.34	1.12	0.05	1.17

demonstrates the estimated scores. A score of 1–2.5 is considered “poor performance,” while a score of 2.6–5 is considered “average performance.” Scores between 5–7.5 are considered “moderate performance,” while scores between 7.6–9 are considered “good performance.” Scores ranging from 9.1 to 10 are considered “excellent performance.”

3. Results

The figure presented as Figure 2 displays the performance of Pakistan’s distribution sector assessed by an index for a period of 22 years. The index is composed of two sub-indices, the “Losses-reliability-Quality-Safety-Index” and the “Recovery-Index.” The efficiency index showed a similar trend as the “Loss-reliability-Quality-Safety-Index,” with the highest scores observed in 2012, followed by the lowest scores in 2021, for both indices. The performance index dropped by 81% between 2012 and 2022, as stated in Table 7.

Over the study period, the efficiency index demonstrated moderate improvement from a poor range in 2006–2009 to an average range, which remained stable between 2010 and 2015 and peaked in 2012 with a score of eight. Nonetheless, the performance index declined significantly between 2016 and 2018, and the trend continued, causing the performance to fall back into the poor range in 2021.

In addition to the “Losses-reliability-Quality-Safety-Index,” the “Recovery-Index” also played a crucial role in assessing the performance of Pakistan’s distribution sector. Although it only had a weightage of 13% in the overall index, it was found to have a significant impact on the sector’s poor performance. The “Recovery-Index” consistently remained in the poor range throughout the entire study period, indicating that the distribution sector struggled to recover its costs efficiently.

During the initial years of the study period from 2001 to 2009, the “Recovery-Index” remained stable without any significant improvement or decline in its score. However, in the later years, it showed an increase in its score, which peaked in 2017. Despite this improvement, the “Recovery-Index” experienced a continuous decline in its score until the end of the study period in 2022.

These findings highlight the need for the distribution sector to focus on improving its cost recovery mechanism to improve its overall performance. Efficient cost recovery mechanisms can lead to an improvement in the sector’s financial stability and help it provide better services to its customers.

Despite encountering various challenges, there have been noteworthy positive developments in specific areas over the last ten years [15,29,30]. For instance, there were reductions in distribution losses by 25%, SAIDI by 18%, load shedding by 65%, faults by 30%, pending connections by 74%, and deaths by 12%. Furthermore, recoveries remained stable, while SAIFI increased by 42%, and the number of complaints increased by 65%.

Nevertheless, despite certain indicators exhibiting progress, their values also showed fluctuations, as indicated by their standard deviation, range values, and variance coefficient. For instance, the narrow range of values for the percentage of power distribution losses over the study period (approximately 2.41 units around the mean) suggests that the power distribution system has not made significant improvements over the past two decades. The lack of progress appears to be due to various factors, including issues with reliability and safety, increased losses, outdated operations, weak institutions, and poor governance. Consequently, the performance index has deteriorated, dropping into the poor range.

In Pakistan, distribution loss is classified into two categories: technical and non-technical. Technical losses are related to the design of the distribution system and can be addressed through innovations, planning, and strict governance [1]. Non-technical losses, on the other hand, are caused by external actions on the power system. As per NEPRA’s findings, IESCO has demonstrated the highest level of efficiency in terms of average losses over the past five years, with losses amounting to 8.682% [10]. Conversely, PESCO has been identified as the least efficient performer, experiencing losses of 37.806%. It should be noted that the average losses across all DISCOs during the last five years have reached 20.27% (Figure 3), which remained relatively high compared to the recommended threshold of 10% for developing nations, as suggested by the relevant literature [22]. Despite NEPRA’s emphasis on achieving a single-digit loss target for distribution companies (DISCOs), the average losses have remained high, indicating the unmet nature of the set target [10]. To address this issue, a reliable metering and recording system is essential to identify unaccounted electricity and technical issues. In the fiscal year 2020–2021, some DISCOs (GEPCO, FESCO, MEPCO, IESCO, LESCO, and K-Electric) met their NEPRA targets, while others (PESCO, QESCO, SEPCO, and HESCO) performed poorly, resulting in an increase in circular debt [11].

Financial losses due to DISCO rates, faulty networks, theft, and poor energy accounting accounted for almost a quarter of the electricity generated lost [4]. With regard to recoveries, their ratios have been poor, highlighting the managerial shortcomings of DISCOs (Table 2) [5]. NEPRA has designated “Recoveries” as a required component for DISCOs to achieve a 100% rate. In 2020–2021, a loss of PKR 160 billion was incurred, and in 2021-22, a loss of PKR 39.4 billion was reported based on DISCO rates [32]. According to NEPRA, the average recoveries of ten different DISCOs in the last five years indicates that most companies have a high recovery rate, with K-Electric having the highest recovery rate of 93.436%, followed by MEPCO at 99.406% [10] (Figure 4). However, QESCO and SEPCO have significantly lower recovery rates of 45.26% and 62%, respectively. Notably, the average recovery ratio for all ten DISCOs was 85.83%. This suggests that most DISCOs in Pakistan were not able to recover a significant portion of their electricity bills from customers. However, it is also important to note that QESCO and SEPCO’s low recovery rates may indicate issues with their billing systems or customer payment behavior. Overall, this suggests that fundamental governance issues in DISCOs have contributed to operational and commercial inefficiencies, including incorrect energy billing and regulatory procedures. Therefore, for sustainable growth, DISCOs must adopt governance as a corporate entity.

The dependability and safety of the distribution network are essential considerations, encompassing factors such as the number of fatalities per year, SAIFI, SAIDI, and losses. Despite reports of deaths resulting from infrastructure problems and citizen awareness in Pakistan from 2020–2022, none of the DISCOs managed to achieve the necessary benchmark for SAIFI and SAIDI during 2013–2021, with an average SAIFI value of 92.69 and an average SAIDI value of 720.9 min, rather than the expected 14 min [33]. IESCO’s reported SAIFI and SAIDI values in 2020–2021 seem to be unrealistic, and some DISCOs exceeded SAIFI targets, while others fell short of SAIDI targets between 2016–2021 [32].

Load shedding has been a major concern in the electricity sector for the last two decades, resulting from insufficient capacity, low power plant efficiency, transmission and distribution losses, and maintenance and upgrading problems. While power generation capacity has risen in the last decade, load shedding decreased only by 50% from 2011–2020, and DISCOs carried out planned and forced outages [10,34]. Despite a reduction in load shedding, complete elimination of the issue has not been achieved, as DISCOs have faced institutional, financial, and governance challenges during the research period. NEPRA has closely monitored the situation and urged DISCOs to minimize excessive load shedding in urban areas. However, it is evident that the average daily load-shedding hours for all listed companies over the past five years amounted to 2.97 h (Figure 5). It is noteworthy that none of the listed companies have reported zero load-shedding hours. Moreover, QESCO has the highest recorded load-shedding hours with 7.686 h per day, whereas FESCO and MEPCO have the lowest load-shedding hours, standing at 0.412 and 0.662 h per day, respectively. These findings offer valuable insights into the current status of load shedding in Pakistan.

Faults per kilometer were also utilized to assess DISCOs’ dependability, with NEPRA suggesting a limit of one fault per kilometer throughout the distribution network. However, several DISCOs (FESCO, SEPCO, GEPCO, IESCO, LESCO, and K-Electric) exceeded this limit, with MEPCO reporting 60 faults per km, highlighting the ambiguity of the supply infrastructure. In conclusion, the study period revealed that the reliability of DISCOs remained a significant issue, with DISCOs struggling to identify fault points in the distribution network.

Customer service plays a critical role in the development of DISCOs as consumers are the lifeblood of these businesses. Social media has made it easier for customers to share their experiences with businesses, and resolving customer complaints in a timely manner is essential to ensure customer satisfaction. To measure the quality of DISCOs, two indicators are used, namely “% pending connections” and “complaints per day.” The failure to provide additional connections to eligible consumers within the specified time frame is attributed to delays in the provision of meters, wires, and power poles. During 2018–2022, QESCO, GEPCO, FESCO, and K-Electric failed to meet their targets, while PESCO, IESCO, LESCO, MEPCO, HESCO, and SEPCO were closer to the 95% target [32]. Failure to provide new connections within the specified time frame has led to a significant financial loss, which increased from PKR 1.4 billion to 2.5 billion from June to December 2021, according to the Ministry of Finance [35].

In the electricity sector, customer complaints are a crucial metric for monitoring the quality of customer service. Based on data from the past five years, the average number of complaints received by electricity distribution companies (DISCOs) in Pakistan is 524,414 (Figure 6). However, there are variations among different companies. Notably, LESCO has the highest average number of complaints, reaching 1,724,188, while SEPCO has the lowest average number of complaints, with only 14,539. DISCOs such as LESCO, MEPCO, and K-Electric, which lack effective mechanisms for handling complaints, were unable to address customer grievances satisfactorily [29,36]. NEPRA has expressed concerns regarding the accuracy of data reported by certain DISCOs, including SEPCO. NEPRA has emphasized the need for DISCOs to establish robust procedures for addressing customer complaints, as well as focus on infrastructure development and inventory management [33].

NEPRA has established safety codes that distribution companies (DISCOs) are required to adhere to in order to ensure the safety and well-being of both their employees and the general public [32]. However, there has been an average of 17.14 deaths per year across all electricity companies due to non-compliance with these safety codes (Figure 7). Among them, K-Electric has recorded the highest average number of deaths per year, amounting to 34.6, whereas QESCO has the lowest average number of deaths per year, totaling 6.8. Unfortunately, the number of fatalities increased to 176 in the year 2020–2021, encompassing both employees and the general public [37]. This suggests that front-line staff across DISCOs were poorly trained and equipped, with an increase in cases during the monsoon season due to the uncovering of power infrastructure and faulty indoor electrical systems, resulting in loss of life [38]. Moreover, DISCOs’ training and facilitation programs have been found to be unsuccessful [39]. There was also a lack of awareness among citizens on how to deal with natural disasters and hazards like electrocution. Therefore, it is crucial for DISCOs to improve their training and safety measures to prevent further loss of life and increase public awareness of safety measures during disasters.

4. Discussion and Policy Implications

The index-based performance evaluation of DISCOs in the period under investigation revealed that the objectives of power sector reforms were not met in letter and spirit, although more than PKR 350 Billion have been allowed to DISCOs in the past seven years for a reduction in losses, network improvement, and customer facilitation [40]. It has been observed that the existing setup has not been able to deliver under the given scenario. Therefore, an estimated USD 17–20 billion [35] is needed to reform the distribution sector and bring about improvements. The following proposals have been put forward for future improvements:

4.1. Governance

Power distribution utilities in Pakistan have faced governance issues for several decades, including political interference, lack of autonomy, inadequate human resources, and financial mismanagement [41]. These issues have had a negative impact on the overall performance of the sector, leading to low reliability and high losses. Several power distribution utilities in Pakistan have faced governance issues related to the points mentioned earlier. For example, the Lahore Electric Supply Company (LESCO) has faced political interference, with the government dissolving the LESCO Board of Directors in 2019 and replacing it with a new board that was criticized for being appointed without following proper procedures [42]. Similarly, K-Electric has faced a lack of autonomy, with decisions often influenced by external factors. In 2012, the government took control of K-Electric and appointed a new board that was criticized for being a political move that did not address the underlying issues facing the utility.

The Islamabad Electric Supply Company (IESCO) has struggled with a lack of skilled human resources, particularly in technical positions [42]. In 2018, IESCO was criticized for failing to hire enough technical staff to maintain and upgrade the power distribution network. The Peshawar Electric Supply Company (PESCO) has faced financial mismanagement, with insufficient investment in infrastructure and maintenance [43]. In 2019, PESCO was criticized for a lack of investment in new transformers, which resulted in power outages and low reliability. Finally, the Quetta Electric Supply Company (QESCO) has struggled with high levels of theft and losses, particularly in rural areas [44]. In 2020, QESCO was criticized for failing to take action against illegal connections and meter tampering, which contributed to the utility’s financial losses.

Addressing these governance issues will require a concerted effort from the government, utilities, and other stakeholders to improve the governance structure and practices in the power distribution sector in Pakistan. The governance issues faced by power distribution utilities are not unique to Pakistan, and other countries have also struggled with similar challenges. For example, in India, power distribution utilities have faced governance issues such as political interference, lack of autonomy, inadequate human resources, financial mismanagement, and high levels of theft and losses [45]. To address these issues, the Indian government has introduced reforms such as the Ujwal DISCOM Assurance Yojana (UDAY), which aims to improve the financial health of power distribution utilities by reducing their debt burden and improving their operational efficiency.

Similarly, in Nigeria, power distribution utilities have faced governance issues such as political interference, lack of autonomy, inadequate human resources, and financial mismanagement [46]. The Nigerian government has introduced reforms such as the Power Sector Recovery Program (PSRP), which aims to address these issues by improving the governance structure of the power sector, strengthening the regulatory framework, and promoting private sector participation.

In South Africa, power distribution utilities have faced governance issues such as political interference, lack of autonomy, inadequate human resources, financial mismanagement, and high levels of theft and losses [47]. To address these issues, the South African government has introduced reforms such as the Electricity Regulation Act, which aims to promote competition and investment in the power sector, and the Municipal Electricity Distribution Company (MEDCo), which aims to improve the governance and performance of municipal power distribution utilities.

To improve governance in the power distribution sector of Pakistan, a multi-stakeholder approach involving the government, utilities, regulators, and other stakeholders is necessary. This approach has been successfully implemented in other countries such as India, Nigeria, and South Africa to promote transparency, accountability, and good governance practices.

4.2. Market Structure

Pakistan is currently taking steps to open up its energy market to competition as a part of its efforts to reform the power sector [48]. The government has been working on institutional arrangements and regulatory frameworks to facilitate the entry of new players in the market, increase competition, and improve efficiency [49]. One of the key reforms is the introduction of the Competitive Trading Bilateral Contract Market (CTBCM) in Pakistan [50]. The CTBCM will allow market participants to buy and sell electricity directly through bilateral contracts without the need for intermediaries. This will facilitate the entry of new players in the market and increase competition.

Another initiative is the establishment of the Independent Power Producers (IPP) Office, which will act as a one-stop-shop for investors and developers looking to enter the power sector [51]. The IPP Office will provide a streamlined process for project development, including approvals, permits, and licenses, to promote investment and competition in the sector. The government is also working on the establishment of a new regulatory authority, the Pakistan Energy Regulatory Authority (PERA), which will have the power to regulate the power sector in a more effective and efficient manner [44]. PERA will be responsible for promoting competition, protecting consumer interests, and ensuring compliance with regulations and standards.

These initiatives demonstrate Pakistan’s commitment to reforming the power sector and improving the energy market. The opening of the energy market to competition is expected to attract investment, increase efficiency, and improve the overall performance of the power sector in Pakistan in the same way as other countries that have opened their energy markets to competition:

• United Kingdom: The UK has a competitive energy market that allows customers to choose their energy supplier [52]. This was made possible through the privatization of the energy industry in the 1980s and the establishment of an independent regulator, Ofgem.
• United States: Many states in the US have deregulated their energy markets, allowing customers to choose their electricity and natural gas suppliers [53]. This has led to increased competition, lower prices, and more innovation in the industry.
• Mexico: Mexico has recently opened its energy market to competition following the passage of energy reform legislation in 2013 [54]. The reforms aimed to increase competition and investment in the energy sector and have led to the entry of new players in the market.
• Norway: Norway has a partially deregulated energy market, which allows customers to choose their electricity supplier [55]. The market is regulated by the Norwegian Water Resources and Energy Directorate, which ensures that prices are fair and that the market is competitive.

4.3. Recoveries Enhancement

The unevenness in the retrieval of dues and outstanding payments among various DISCOs is not exclusive to Pakistan’s energy industry. Such imbalances have resulted in significant repercussions on the financial viability and operational efficiency of the power sector [56]. For example, in the last five years, the average recoveries in QESCO were 46%, 63% in SEPCO, and 65% in HESCO, and the rest were at more than 92% on average [32]. Therefore, it is essential to tackle these issues to ensure a dependable and reasonably priced electricity supply. In order to achieve this objective, Pakistan could take cues from other countries that have implemented various initiatives to improve their recovery and arrears management. For example, in Ghana, the Electricity Company of Ghana (ECG) has struggled with high levels of non-payment by consumers, resulting in significant arrears and difficulties in financing new infrastructure projects [57]. The ECG has implemented several measures to improve recoveries, including the installation of prepaid meters and a customer billing and payment system.

Similarly, in Nigeria, distribution companies (DISCOs) have faced significant arrears and non-payment by consumers [46]. This has led to difficulties in financing new infrastructure projects and has resulted in frequent power outages. The Nigerian government has implemented several measures to address this issue, including the introduction of the Meter Asset Provider (MAP) program, which aims to provide prepaid meters to consumers and improve revenue collection for the DISCOs [58].

In South Africa, the power utility Eskom has struggled with significant arrears and non-payment by municipalities, resulting in difficulties in financing new infrastructure projects and maintaining the existing power infrastructure [59]. Eskom has implemented several measures to address this issue, including the introduction of a revenue protection unit to improve collections and the installation of prepaid meters.

Moreover, DISCOs must choose the best strategy, message, and channel to reach out to a consumer. The instruction of taking prompt action when a consumer’s arrears exceed the prescribed limits must be strictly adhered to. In turn, it may provide a better probability of securing a commitment to pay while keeping a positive consumer perception. In the event of a default, re-segmenting consumers and determining the optimum exit path for each of them is crucial, leaning toward fair results and the best customer experience. Service suspension, when unavoidable, should be managed; for example, DISCOs must release the list of defaulters to the media and their websites and reconnect once dues are paid.

It is observed that DISCOs (QESCO, GEPCO, SEPCO, IESCO, and HESCO) were understaffed [3]. They may take the assistance of outside agencies to fill the void. Outsourcing, for example, may be used to identify high-risk consumers as early as feasible and provide them with choices to minimize late payments [60]. Furthermore, outsourcing may aid in reducing administrative errors in order to avoid late payments and non-payment. Keeping invoicing errors to a minimum would improve the customer experience while also lowering a DISCO’s bad debt. DISCOs should roll out electronic bills to their whole customer base and regularly monitor the end-to-end invoicing process to eliminate leakages to further boost recoveries. Spot billing and collection, as well as check drop and pre-payment facilities, should be encouraged and made available.

4.4. Conservation Provision

During peak hours, Pakistan’s distribution system can reliably handle loads of only 22,000 MW [61]. In other words, the distribution network is a more restrictive constraint than the generation capacity. This point of view is aided by the fact that the distribution network is particularly vulnerable to failure when public pressure drives DISCOs to supply more power. In May 2020, for example, distribution experienced repeated outages as the load on distribution networks exceeded 22,000 MW for an extended period of time [30]. During this time, distribution constraints accounted for nearly 80% of load management. Any increase in generation capacity (or even the payoff of the circular debt) will not be able to alleviate load management on a long-term basis unless the existing distribution network is upgraded [62].

However, financial constraints would arise for distribution upgradation. Instead, conservation measures must be adopted to curb the demand, just like countries have adopted. For example, AusNet Services (Australia) offers rebates and incentives for customers who install energy-efficient appliances or participate in demand response programs [63]. After the Fukushima nuclear disaster in 2011, Japan implemented aggressive energy conservation measures to reduce electricity demand [64]. These measures included a national “Cool Biz” campaign, which encouraged workers to dress casually to reduce air conditioning usage, and a “Super Cool Biz” campaign, which encouraged even more casual dress in the summer months. The country also implemented a policy of mandatory energy audits for large businesses and established subsidies for the installation of energy-efficient appliances [65].

In order to reduce electricity demand and encourage sustainable energy consumption, DISCOs like K-Electric, HESCO, LESCO, and MEPCO must vigorously implement conservation measures. They may need to engage with appliance manufacturers to raise awareness and promote energy-efficient products. Furthermore, the government should consider reducing taxes on energy-efficient appliances, such as those with energy star ratings and LED-based lighting, and provide substantial subsidies for them.

Furthermore, there is a need to revise building codes in Pakistan to ensure that new residential, office, commercial, and industrial buildings are constructed with greater energy efficiency in mind [66]. To achieve this, restrictions may be placed on the use of air conditioners during peak hours, similar to the measures implemented in Malaysia [67]. The implementation of tax and credit incentives for the industrial sector can encourage businesses to invest in conservation programs that promote energy efficiency and reduce their overall energy consumption. For example, in the United States, businesses can receive a tax deduction of up to USD 1.80 per square foot for building improvements that save energy in lighting, heating, cooling, ventilation, and hot water systems [68]. Similarly, in India, the government offers a range of incentives for industries that implement energy-saving measures, including a 10–15% subsidy on the capital cost of energy-saving projects, accelerated depreciation on energy-saving assets, and concessional financing for energy efficiency projects [69]. Lastly, GoP must take an active role for longevity to have a meaningful impact through media and advertising campaigns to create awareness among users.

4.5. Distribution Losses

Over the past five years, DISCOs such as PESCO, QESCO, SEPCO, and HESCO have reported average losses of 37%, 25%, 36%, and 28%, respectively [32]. To address these losses, a long-term strategy is needed at the DISCO level, particularly for these specific DISCOs. While targets may differ among the DISCOs, a 50% reduction in losses within a five-year time frame should be considered a benchmark for all DISCOs [70].

Furthermore, the utilization of Distributed Generation (DG) can be considered, which refers to the use of small-scale power generation units located near the point of consumption, such as solar panels, wind turbines, or small gas-fired generators. By implementing DGs, power system losses can be reduced, leading to improved efficiency and reliability of the system [71]. The Load Concentration Factor (LCF) and placement of capacitors method can be utilized to identify the optimal locations for DG placement based on the load concentration factor [72,73]. Another option to consider is the fragmentation of DISCOs into smaller ones. In this regard, there have been successful examples such as a DISCO (known as DISCOM in India) located in the state of Andhra Pradesh, which was able to reduce its losses from 12.98% to 10.68% within four years and eventually to six percent [74].

There are indeed multiple methods available to curtail distribution losses in a power system, considering that such losses can lead to a substantial amount of energy waste. Pakistan can utilize some of these measures (below) utilized in various countries to improve their power system’s efficiency and lower the amount of energy that was lost during distribution:

• Replacing old, bare wire distribution lines with PVC-coated power cables: This can help to minimize leakage and reduce the losses that occur as a result of electrical resistance in the distribution lines. For example, in India, the government has initiated a program called the “Integrated Power Development Scheme” (IPDS) to improve the power distribution infrastructure, which includes replacing old distribution lines with PVC-coated power cables [69].
• Running the entire distribution system underground: This can help to protect the system from weather-related damage, vandalism, and theft, which can contribute to losses. For example, in Nigeria, the government has embarked on an initiative to move the country’s power distribution system underground to protect it from vandalism and theft [46]. For example, the Enugu Electricity Distribution Company (EEDC) has recently started implementing an underground cabling project in some areas.
• Upgrading transformers: Replacing old transformers with newer, more efficient ones can reduce losses by decreasing the amount of power that is wasted as heat. In this regard, the government of Bangladesh has launched a project to upgrade the country’s distribution transformers to improve their efficiency and reduce losses [48]. The project aims to replace around 1.3 million transformers with more efficient ones by 2023.
• Implementing power factor correction: Power factor correction involves improving the efficiency of the electrical system by reducing the reactive power that is generated and improving the power factor. This can help to reduce losses and improve the overall efficiency of the system. One example is from Egypt, where the government has implemented a program called “Egypt Energy Efficiency and Renewable Energy” (EEERE) to improve the efficiency of the country’s power systems [75]. The program includes implementing power factor correction measures in various industrial and commercial sectors.
• Implementing voltage regulation: Voltage regulation involves keeping the voltage levels within the required range, which can help to reduce losses by preventing over-voltage or under-voltage conditions. An example of this is Brazil, where the government has implemented a program called “Prodist” to modernize the country’s power distribution system, which includes implementing measures to regulate voltage levels [76]. The program aims to reduce losses by improving the efficiency of the distribution system and reducing the amount of energy that is wasted due to over-voltage or under-voltage conditions. As part of this program, the government has invested in new equipment such as voltage regulators, capacitors, and transformers to improve the voltage profile of the distribution network. The program has helped to reduce losses and improve the reliability of the power supply in Brazil.

While the options mentioned earlier are effective ways to reduce distribution losses in a power system, there are other measures that can also be implemented to achieve similar outcomes. For example:

• Use of high-efficiency distribution transformers: Replacing old, inefficient transformers with high-efficiency models can help reduce distribution losses [77].
• Reduction of technical losses: Improving the power factor of the distribution network and reducing system voltage drops can also help to reduce technical losses in the power system [78].
• Energy auditing and monitoring: Conducting regular energy audits and monitoring the distribution system can help identify areas of inefficiency and opportunities for improvement [79].

By implementing a combination of these measures, along with those mentioned earlier, significant energy savings and cost reductions can be achieved in a power system in Pakistan.

4.6. Safety

In many developing countries, safety measures and regulations in the power sector are often inadequate, leading to a higher risk of accidents and fatalities [80]. For instance, in India, there have been several incidents of electrocution and other safety-related accidents due to poor maintenance of distribution infrastructure and inadequate safety measures [69]. In response, the government has introduced several measures to improve safety, such as the mandatory use of insulated wires and cables and the installation of ground fault circuit interrupters (GFCIs) in homes. Similarly, in Pakistan, there have been reports of safety violations and accidents in the power sector [10]. As a result, the average number of deaths per year resulting from accidents or incidents involving both groups combined is 17, as indicated in Table 2. The DISCOs, such as K-electric, averaged 34 deaths per year, whereas PESCO, IESCO, and HESCO averaged 22 deaths per year. The rest of the DISCOs remain at 12 deaths per year on average.

Therefore, Pakistan needs to prioritize the implementation of stricter safety measures and regulations in the power sector to prevent accidents and fatalities. This requires a comprehensive approach that involves government regulation, industry compliance, and proper training and equipment for workers. In this context, examples from the developed nations can be followed. For example, in the United States, the Occupational Safety and Health Administration (OSHA) sets and enforces safety standards for workers in the electricity generation, transmission, and distribution industries [81]. OSHA’s standards cover a wide range of safety measures, including fall protection, protective equipment, and electrical safety, among others.

Similarly, in Japan, the Ministry of Health, Labor, and Welfare (MHLW) has established safety guidelines for electrical work, which include requirements for safety equipment, procedures, and training [82]. The guidelines apply to workers in all sectors of the electrical industry, including power generation, transmission, and distribution. Resultantly, according to the International Energy Agency (IEA), the United States had an average of 0.05 deaths per terawatt hour (TWh) of electricity generated between 2011 and 2022 [83]. Japan had an average of 0.01 deaths per TWh during the same period.

In addition, NEPRA should launch investigations and legal proceedings against any DISCOs in order to levy heavy fines [70]. Given the value of human life, NEPRA should direct DISCOs to compensate victim families of public persons for whom it has been held liable with the same amount paid to its employee. Furthermore, DISCOs may be directed to provide employment to the next of kin of public figures who have lost their lives. In addition, DISCOs must formulate the safety manual based on human psychology and behavioral factors [84]. In addition, examples of other countries can also be benchmarked, such as Pacific Gas and Electric Company (PG&E) (USA) has implemented a comprehensive safety program that included regular inspections of equipment, as well as training for employees and customers on how to stay safe around electricity [68]. Another example is Enedis (France) launching a safety awareness campaign aimed at reducing the number of accidents related to electricity and gas [85].

4.7. Load Shedding

In spite of having sufficient generation capacity to fulfill the electricity demand, DISCOs in Pakistan have implemented a policy of load shedding [86]. This load shedding was observed on the feeder level in several locations across Pakistan, citing reasons such as excessive losses, theft, and low recoveries. On average, the load shedding across all DISCOs was reported to be 2.3 h per day, with HESCO and PESCO being the worst affected, with an average between five and seven hours of load shedding per day, respectively [32]. This policy has been deemed unfair to customers who pay their bills and taxes [87]. Therefore, DISCOs must improve their governance and disconnect individual customers who are defaulters or involved in electricity theft.

In this situation, it is recommended that the Government of Pakistan (GoP) instructs NEPRA and DISCOs to prioritize the construction of grid infrastructure. Funding should be allocated as soon as possible, and options for financing from donors should also be pursued [77]. DISCOs must also implement the best engineering practices for designing and developing distribution networks to prevent load shedding in the future [6]. Predictive maintenance should be implemented across all DISCOs to increase efficiency and minimize costs [70]. Predictive maintenance is a tool used to analyze the operational equipment’s status and forecast maintenance requirements to prevent equipment failures [88].

Other causes of load shedding, such as fuel scarcity, must also be addressed. About 60% of power generation in Pakistan depends on fuel imports, and the delay in imports caused generation plants to fail to receive fuel in time for power generation [61]. To reduce the import bill and minimize load shedding, the penetration of distributed generation and off-grid generation should be increased, particularly through renewable energy sources such as photovoltaic energy [17]. Solar technology has become economically feasible and reliable in Pakistan and is the most suitable candidate for large-scale off-grid energy generation [6]. Appropriate subsidies and incentives should be provided to encourage the installation of off-grid renewable resources [89]. In this example, Denmark can be followed as it has implemented a similar system to Germany called the “Renewable Energy Feed-in Tariff Scheme” (REFIT) [90]. This scheme provides financial incentives for the production of renewable energy, including wind, solar, and biomass energy. As a result, Denmark is one of the world leaders in renewable energy production.

4.8. Theft Control

Electricity theft is a significant issue in Pakistan, with reports indicating that an average of 20% of all electricity supplied in the country is stolen [91]. A concrete and scientific solution is required to address this problem. One potential solution for combating electricity theft globally is the implementation of “smart metering” and “prepaid metering,” with prepaid meters being particularly effective [92]. By replacing the existing metering system with prepaid meters, DISCOs can prevent theft. Prepaid meters are equipped with an electrical meter that automatically turns off when the credit or amount expires and is difficult to tamper with [93]. Additionally, addressing theft requires addressing governance systems, including civil rights, penalties, democratic institutions, and accountability. Rules and guidelines should be strengthened to deter electricity theft, and literature should be consulted for technological, institutional, legal, and policy solutions to combat the problem [94]. Enforcement should involve imposing fines on individuals who commit electricity crimes and may be initiated by the utility company with state resources and laws provided by the government [6]. Furthermore, meters should be installed outside the end-premises user’s location, and old naked wire distribution lines should be replaced with “Aerial Bundled Cables” (ABCs), or the entire distribution system should be run underground [11]. K-Electric’s Ujala project provides a successful example of ABC implementation in Karachi [95]. This project involved the installation of new transformers and low-cost connections and the use of ABC, which significantly improved power supply reliability and eliminated illegal connections for nearly 200,000 consumers. Other DISCOs should follow K-Electric’s lead and initiate ABC projects with appropriate funding allocations.

4.9. Quality of Services

In order to improve customer service standards, DISCOs in Pakistan should adopt a service-oriented approach by building infrastructure to interact with consumers through the use of information technology. Other countries have implemented various IT solutions in their power sector to improve customer service. For example, the United States has implemented an automated meter reading system that allows utility companies to monitor the usage of electricity remotely, eliminating the need for manual meter reading [96]. This system also enables customers to track their energy usage in real-time, providing them with better control over their energy consumption.

In Australia, the use of electronic payments has been widely adopted by utility companies, enabling customers to pay their bills online through various payment methods, including credit cards and direct debit [97]. This approach has made the payment process more convenient for customers and has reduced the workload for utility companies.

To improve customer service standards, DISCOs in Pakistan can also benefit from implementing IT solutions such as e-mail bills, electronic payments, automation in meter reading systems, underground cabling, DISCO mobile apps, and other similar technologies. These technologies can enable DISCOs to become paperless, reduce personal interaction with customers, and enable timely responses to consumer complaints. However, this goal would require the construction of infrastructure that reduces the need for personal interaction with customers and facilitates prompt resolution of consumer complaints [77].

Another option is to build more customer facilitation centers across Pakistan to improve service quality. Additionally, to bring professionalism and good engineering practices to DISCOs, higher officers must supervise and regularly check the quality of work of their line staff [78]. Moreover, staff at complaint centers require extensive training programs to develop the skills necessary for processing and handling complaints. Failure to provide such training may lead to dissatisfied customers who may not pay their bills on time or may engage in electricity theft.

This study proposes a mixed approach to enhance the efficiency of the power sector, which includes developing indices and providing comprehensive policy recommendations based on findings. Previous studies, such as [48,69], focused on discussing the reforms and proposing conceptual frameworks for the power distribution sector in India and Bangladesh. However, the challenges in the power distribution sector in these countries are similar [98,99], and the methodology and indicators used in this investigation can be extended to utilities in those countries for further efficiency improvements.

The methodology used in this study was tested for sensitivity and reliability, as emphasized in [100,101], and the results were found to be robust (Figure 8). However, there are limitations to this methodology, such as the difficulty in interpreting the first few principal components and the assumption of linear relationships between variables in PCA. Therefore, future studies can explore other techniques and weightage schemes, such as distance to reference and analytical hierarchy process, to enhance the evaluation process and policy recommendations for DISCOs.

5. Conclusions

This study utilized an integrated framework to evaluate the performance of Pakistan’s power distribution industry and identified several factors contributing to its decline from a moderate to poor range. These factors include governance issues, uneven retrieval of dues, high losses, electricity theft, and safety concerns resulting in an average of 17 deaths per year. To improve the sector’s performance, the study recommends a range of measures, including underground distribution systems, energy auditing, and monitoring, compliance with safety standards set by OSHA, addressing fuel scarcity to reduce load shedding, adoption of “smart metering” and “prepaid metering,” and infrastructure development for enhanced consumer interaction through information technology. Moreover, a customer-centric approach should be adopted by DISCOs to elevate customer service standards. Furthermore, there is a need for additional research to establish a framework enabling DISCOs to incorporate renewable energy capacity into their systems. Additionally, Pakistan’s efforts to introduce competition in its energy market as part of power sector reforms should be continued.

It is worth highlighting that neither the DISCOs nor NEPRA have incorporated alternative renewable energies (AREs) into their power distribution plans, as indicated in previous studies [102,103]. The National Grid (UK) serves as an example, having made significant investments in upgrading its electricity transmission network to integrate renewable energy sources [104]. Thus, further research is necessary to develop a framework for DISCOs to adopt an integrated approach, address underlying inefficiencies, and leverage new technologies to mitigate losses and infrastructure constraints.

Future research endeavors should venture into exploring alternative techniques that go beyond Principal Component Analysis (PCA) and also consider different weighting schemes. While PCA is a commonly used dimension reduction technique, it has certain limitations in terms of linearity and sensitivity. Moreover, in addition to exploring alternative techniques, considering different weighting schemes can contribute to a more comprehensive evaluation process. Weighting schemes determine the relative importance assigned to different indicators or dimensions. By adopting different weighting schemes, researchers can assess the sensitivity of the evaluation results to variations in the importance assigned to various factors.

An innovative idea for future research, building upon the aforementioned context, could involve exploring the potential of machine learning algorithms to automate the weight estimation process for sectoral indicators in assessing the performance of DISCOs. By integrating machine learning algorithms into this process, researchers can introduce a more objective and data-driven approach, thereby increasing the robustness of index results and reducing potential biases associated with relying solely on expert opinions. This research idea contributes to bridging the existing literature gap by offering insights into the effectiveness and suitability of machine learning techniques for quantifying and strategizing the performance of distribution utilities using sectoral indicators and dimensions.