New Accountability Approach: Utilising Dynamic Zero-Waste Baselines to Mitigate Water Wastage in Gold Mines

Abstract

The South African gold mining industry requires complex water reticulation systems to deliver chilled water to underground production areas. However, chilled- and service-water wastage, including leaks and misuse, contribute to approximately 50% of the total chilled-water demand. The current inefficiency detection methods rely on broad, infrequent, and labour-intensive work, focusing only on identifying and quantifying wastages without comprehensive mitigation strategies. This study aimed to develop a novel accountability framework employing dynamic zero-waste baselines to identify and address inefficiencies closer to active working areas. The proposed method incorporates four key components—define, assess, execute, and communicate—into an accountability system to monitor performance and ensure sustainable improvements. The integration of dynamic zero-waste baselines within this accountability framework will ensure faster and more accurate inefficiency detection and, more importantly, the mitigation thereof, significantly reducing water wastage. This study successfully reduced the daily water wastage, with an annual energy cost benefit of approximately USD 1.6 million (ZAR 28.7 million). The successful implementation of this method met all the research objectives, confirming its effectiveness.

1. Introduction

1.1. Background

Chilled water plays a vital role in South African gold mines. Its utility extends to underground air cooling and various mining activities, including the cooling of rock drills, cleaning/sweeping, and dust suppression [1,2]. Chilled-water pipelines are highly intricate and complex, supplying the end users with chilled water 4 km below the earth’s surface, with these pipelines reaching lengths of over 15 km [3,4,5]. Presently, the water reticulation system, inclusive of pumping and refrigeration, constitutes approximately 39% of the overall energy consumption of a typical South African gold mine [6].

As mining operations progressively extend to greater depths in pursuit of new gold reefs, operational costs increase, further intensified by escalating electricity tariffs [7]. Consequently, there is an ongoing need to curtail the operational costs of South African gold mining operations [8].

Considering the complex nature of water reticulation systems, together with the harsh underground conditions, inefficiencies in the form of leaks and misuse are prone to occur [9]. Previous studies have shown that up to 50% of the total water used can be attributed to wastages [10].

Due to the intricate nature and scale of water reticulation systems, the ongoing challenge lies in the identification and quantification of the inefficiencies [6]. Addressing the inefficiencies is essential for reducing operational costs and extending the life-of-mine (LOM) of gold mining operations. However, before the inefficiencies can be mitigated, they need to be identified.

The current inefficiency detection methods include intensive underground audits, step testing, zero-waste baselines, hardware and software methods, benchmarking, historic baselines, and control valves [9,11,12,13,14,15,16,17]. These methods can identify the inefficiencies per mining level, still requiring labour-intensive measures to localise the issues.

A recent paper published by Le Roux [18] compared several wastage identification techniques to determine the effectiveness of each. Le Roux [18] considered techniques such as benchmarking, historic baselines, zero-waste baselines, leak detection, and control valves. Each technique was applied to a single mutual case study to provide a comparative platform. The paper concluded that the most accurate representation of wastages is the zero-waste baseline method, as all other techniques neglect some elements of wastage.

1.2. Zero-Waste Baseline Methods

A zero-waste baseline is described as the theoretical minimum consumption considering an idealised scenario, devoid of any wastages in a system [6,11]. This minimum consumption signifies the operational requirements achievable at 100% efficiency [19]. The utilisation of a zero-waste baseline is commonly employed to systematically identify and quantify wastages within a given system [14,20].

As mentioned, research has been conducted regarding the utilisation of zero-waste baselines, with a specific focus on mining operations [6,11,12,21,22,23]. The current zero-waste baselines have been developed per mining level. This means that inefficiencies can only be identified per mining level, still requiring labour-intensive underground audits to pinpoint the inefficiencies. Furthermore, the current zero-waste baselines have been developed using fixed timetables and fixed demands for mining equipment obtained from their specification sheets (using design specifications).

One study’s unique approach included the localisation of inefficiencies in closer proximity to the working areas, namely that of Jordaan et al. [6].

Jordaan [6] made use of dynamic zero-waste baselines to effectively identify inefficiencies. Jordaan showed that most inefficiencies tend to occur closer to the stopes/working areas where the gold-bearing ore is mined. However, service water monitoring is currently only performed up to the half-level splits (measuring all the crosscuts on a half level together), which is not sufficient when considering the complex layout of underground mining operations. Jordaan installed flow and pressure meters in each working area to closely monitor the service water consumption in closer proximity to the working areas. Once the actual flow profiles for each working area were established, Jordaan developed a dynamic zero-waste baseline using the Missing Persons Locator (MPL) system’s data, mining schedules, and correlation models, derived for each water user to introduce variability based on the underground pressures of specific areas. This approach ensured the quicker and more precise identification of inefficiencies, allowing for them to be pinpointed to specific working areas. These inefficiencies could then be addressed and continuously monitored to ensure sustainability.

The use of zero-waste baselines to identify inefficiencies in the mining industry is not a new concept. However, no studies could be found where a method was developed to actively ensure that the identified inefficiencies were mitigated and monitored for sustainability. Although it remains important to identify inefficiencies, it is meaningless if they are not addressed and managed. Responsibility allocation should be incorporated to assign the identified inefficiencies to certain personnel who will be held accountable for their mitigation [6].

1.3. Accountability Review

The term accountability is often regarded as a multifaceted and evolving concept with multiple definitions [24,25]. The literature on accountability confirms that it is rather disconnected, as authors have set out to produce their own definition depending on the context and purpose for which it is sought [26,27]. It has been used as a synonym for good governance, transparency, equity, democracy, efficiency, responsiveness, responsibility, and integrity [26]. However, the concept of accountability can be simply defined as the process of being held responsible for one’s actions [28,29,30,31]. Individuals and/or organisations are held accountable by reporting to a recognised authority (or authorities) [30,31].

Accountability has been described as the root of a viable social system, as the need for employees to be held accountable for good performance is widely accepted [32]. It has been found that there is a strong link between performance, behaviour, and accountability [33,34]. Consequently, accountability systems play a vital role in organisations by ensuring responsibility for actions, decisions, and performance. The main components of accountability, as identified in the literature, are shown in

Table 1. Accountability components summarised from conducted research.

Component	Description
Clarified expectations	Mero et al. mentioned that outcomes are achieved when expectations are clear [33]. No accountability can be implemented if personnel do not understand what is expected from them. Expectations regarding roles, responsibilities, and performance standards should be clarified together with rewards and penalties [33,35,36].
Improved performance	Brandling et al. found that accountability measures improve employee performance and, by doing so, operational performance [36]. Performance appraisal systems should be implemented to track and monitor performance [37]. These include regular feedback, performance reviews, and goal tracking to allow individuals to identify areas for improvement. This promotes continuous learning and development of organisations.
Transparency and fairness	All actions and information should be made accessible to all employees to ensure transparency [30]. Furthermore, a culture of fairness should be fostered whereby all employees are held to the same standards [38].
Risk management	Assistance with risk mitigation by identifying certain risks and taking proactive measures to prevent them. By establishing accountability towards compliance with regulations, ethical standards, and best practice, organisations can reduce the likelihood of costly mistakes, legal liabilities, and reputational damage [30,36].
Goal alignment	Organisational and individual goals are aligned to ensure that everyone is working towards common objectives [33,39]. Organisations can maximise their efficiency by linking individual performance to their priorities and strategies [37].
Employee engagement and motivation	Employees are motivated as they are assigned responsibilities and receive recognition for their contributions [33,35]. When an accountability culture is fostered, employees hold each other and themselves accountable as improvement efforts are duly recognised, leading to engagement, motivation, and commitment to reaching organisational targets [36].

.

The research has highlighted four main methods of accountability, which are as follows [35,36]:

• Self-accountability: This refers to individuals taking responsibility for themselves. This overlaps with components of self-criticism, self-monitoring, and self-management. This is coupled with one’s upbringing, values, and motivation to have a desire for achievement and inner satisfaction (an individual’s sense of responsibility) [36].
• Peer-to-peer accountability: This form of accountability occurs between individuals at the same hierarchical level within an organisation. It involves reciprocal relationships and a sense of accountability towards one another, where individuals hold each other responsible for their actions and decisions. This type of accountability relies on peer interactions and mutual trust rather than hierarchical oversight. However, individuals may prioritise their personal reputation when being held accountable. Overall, while peer-to-peer accountability promotes collaboration and empowerment, it may also introduce ambiguity, bias, and conflicts in accountability assessments and decision-making processes.
• Managerial accountability: Managerial accountability involves individuals facing higher authorities to account for their performance, with hierarchical systems granting managers the power in this process. The research stresses that this mechanism is most effective when employees must face their managers in person to account for their performance. The negative aspects of managerial accountability include stress, when only failures are recognised, and perceptions of excessive control. The positive aspects include the recognition of good performance and showing compassion towards employees, leading to a positive association with accountability.
• Systems accountability: Systems accountability refers to the mechanisms and processes implemented within organisations to monitor, regulate, and ensure accountability for individual actions and performance. These mechanisms often include formal procedures, policies, and systems, such as performance appraisal systems, accounting systems, surveillance mechanisms, and computer monitoring. The goal of systems accountability is to establish clear standards, track performance, and enforce accountability to promote organisational effectiveness and compliance with the goals and objectives. While the significant impact of systems accountability is indisputable, it is essential that performance measurement and monitoring are performed accurately.

Accountability is a well-researched topic and has proven to be a powerful tool to use within organisations. However, accountability in the mining industry remains an unexplored topic. One relevant recent paper by Brandling et al. [36] examined accountability across different organisational levels within a mining company, revealing that the workers, supervisors, and managers perceived accountability differently, and focused on the rules, performance standards, and decision-making authority, respectively. Their study highlighted the importance of communication, role clarification, trust, discipline, and safety, especially in high-risk environments. However, it did not implement a specific accountability method, instead showing how different levels view it.

Pearson et al. [35] addressed this gap by developing a “System of Accountability” model based on interviews with 20 leaders, identifying five key factors: organisational systems, culture, role clarity, leadership, and individual attributes. These factors are critical for enhancing accountability and improving organisational performance. While accountability leads to better performance, mining companies often lack clear success metrics, according to Isacowitz [40], making it harder to measure outcomes effectively.

Developed systems play a crucial part in the development of successful accountability methods. It involves clear communication, role clarification, and transparency, which have all been identified as critical by researchers. While self-accountability and organisational culture are vital, managerial accountability provides the foundation that will ultimately determine the success of the subsequently described accountability types. A hierarchical approach ensures a holistic goal alignment, disciplined decision making, and unifies the actions that contribute to organisational success.

Accountability is a key element to reduce wastages [41]. However, the application of the outlined methods without their coherent adaptation to the economic, social, and demographic limitations associated with the mining industry can be detrimental to workforce morale and organisational profitability [42]. References to an incentive or sanction structure can produce various union-related complications, such as strike actions. The research further highlights that the consequences do not only include incentives or sanctions, but can also include social pressure, where others will know if an expectation has been met, praise or critique, to even having an impact on one’s career or having to answer to a higher authority [43]. It is, therefore, crucial to evaluate the influential parameters to reconstruct existing accountability methods to conform to the mining industry.

1.4. Research Aim and Objectives

The research highlights the need to develop a framework that will ensure that identified inefficiencies are addressed. By establishing clear lines of responsibility and fostering a culture of accountability, mining operations can sustainably avoid unnecessary expenses and optimise resource utilisation.

This aligns with a significant limitation identified in the prior research regarding the absence of a well-defined post-identification method for addressing inefficiencies. While certain studies have incorporated feedback sessions, a substantive gap remains toward establishing a robust method for holding individuals accountable for issues within their designated areas.

Therefore, this paper includes a novel approach where zero-waste baselines will be integrated with an accountability method for mining operations. This combined approach seeks to ensure personnel accountability for wastage within their designated areas, thereby contributing to a reduction in the operational expenditures.

Therefore, the objectives of this paper are as follows:

• The development of an accountability method for mining operations to address costly inefficiencies.
• The incorporation of zero-waste baselines into the method as an accurate performance benchmarking metric.
• Expeditious inefficiency detection and mitigation in comparison to previous methods.
• Ensuring that the developed method avoids labour-related concerns by steering clear of financial incentives.

The method used to address the objectives of this paper is discussed in the following section. The newly proposed method develops dynamic zero-waste baselines for each working area to enable inefficiency detection throughout the day and in closer proximity to working areas where the majority of inefficiencies are found. Due to the localisation of inefficiencies, responsibility allocation is implemented to ensure that the identified issues are mitigated. The continuous monitoring of each working area is implemented, which leads to decreased water wastage and ensures the sustainability of the mitigated inefficiencies.

2. Method

Four key steps are highlighted that should be included in a bespoke accountability model, namely define, assess, execute, and communicate [44]. These four steps are used as the basis of this method. The components described in Table 1 will be incorporated into and addressed with the implementation of the method. The proposed method is displayed in Figure 1.

2.1. Step 1: Define

2.1.1. Clarify Expectations, Goal Alignment, and Transparency

Table 1 addresses the importance of clarified expectations, goal alignment, and transparency within an accountability method. Clear expectations should be established so that employees have a thorough understanding of the success criteria. In this context, success will be represented by the dynamic zero-waste baselines, which will be developed for each individual working area. These baselines will display the best practices for water consumption, serving as the operational standard. The success metrics will dynamically adjust based on the operational conditions and employee activity in each working area. Ensuring that all personnel are aware of the expectations not only promotes transparency, but also aligns with organisational and individual goals [30].

2.1.2. Major Water Consumers

The service water distribution in deep-level gold mines is divided between continuous and intermittent water users [1]. Continuous users, including air-cooling units such as bulk air coolers and mobile cooling units, receive a continuous water supply to maintain underground conditions in compliance with the Health and Safety Act of 1996 [45]. The focus here lies on the intermittent water users, who are the non-continuous water consumers in gold mining operations. The primary intermittent water consumers in gold mining operations include [46] the following:

• Rubber hoses;
• Rockdrills;
• Waterjets;
• Sweeping tools;
• Support tools.

To uphold accountability, it is essential to ensure that performance measurements are accurate, promoting fairness across the board. Conversely, previous researchers have tended to assume a fixed water consumption rate for water-consuming equipment obtained from specification sheets [11,23]. However, this approach overlooks the fact that some equipment, especially those manufactured in-house, may lack specification sheets. Furthermore, the data on equipment specification sheets typically represent the optimal operating conditions, which are determined within controlled environments and may neglect the performance in a deep-level gold mine. Therefore, to address this gap, tests should be conducted on identified water-consuming equipment to accurately determine their actual water consumption for specific operations. By doing this, all the water-consuming equipment will be considered, and their usage accurately accounted for. The tests should involve an underground test bench where water consumption can be evaluated by adjusting the water pressure. Measurements should be taken at different pressure levels using a portable flowmeter and pressure gauge.

The test data should be represented on a scatter plot, and the trendline that best fits the data should be chosen based on the $R^2$ value closest to 1 [6]. This trendline equation will be used to represent the water consumption at various pressures.

2.1.3. Employee Tracking Incorporation

The water consumption of all users has now been determined; however, the allocation throughout the day should be established. The literature indicates that water allocation for zero-waste baselines is performed for fixed periods corresponding to the working shifts of gold mines [6]. This approach may not provide an accurate depiction of water usage. This is due to the fact that water consumption by users may vary throughout shifts, leading to potential inaccuracies in its allocation [11,12,19]. Most of the intermittent water-consuming equipment consists of handheld, manually operated tools used by mining personnel. These tools only consume water when actively used by employees. Therefore, if there are no employees present in a working area, these handheld tools should not consume water.

Water usage during times when employees are not present will be classified as a misuse or inefficiency. Moreover, the operation of these tools requires specialised training to ensure their safe and optimal use. Consequently, if an employee trained to operate specific tools is not present in a given working area, those tools should remain unused. Therefore, having knowledge of the employees’ whereabouts in the working areas enables the allocation of water usage based on the specific equipment they are trained to operate.

To obtain an employee’s presence in a specific working area, an employee tracking system can be used. Employee tracking systems are widely implemented in deep-level gold mines to track employees’ locations in case of emergencies [47]. Employee tracking systems typically make use of a radio frequency identification (RFID) system, integrated into the employees’ headlamps, which transmits unique identification codes intermittently [48,49,50,51]. These codes are linked to the employees’ clock cards, containing personal information, such as the employee name, employee number, operation name, region, zone, and job description.

RFID tag readers, installed at fixed locations throughout a mine, detects the nearby tags and transmits their codes to a central server, enabling the real-time tracking of employees within each tag reader’s coverage zone [52,53]. This system enhances safety and operational efficiency by providing real-time employee location data. The data logged by the RFID tag readers are essential for employee tracking. They will be used to establish personnel movements, which directly correlate to water consumption patterns. Inefficiencies will be pinpointed, as water usage will be associated with specific activities and used as part of a dynamic zero-waste baseline development. The logical map of the process discussed is shown in Figure 2.

2.1.4. Dynamic Zero-Waste Baseline Development

The zero-waste baselines will be established by integrating the above-mentioned factors. An area’s measured water pressure will be inserted into the developed formula to determine the specific equipment’s water consumption. The personnel activity will be used to determine the type of water user at any time of the day and assign the corresponding water-consuming tool to them. Doing this will ensure a 24 h benchmark profile that will be measured against the actual measured reading. These baselines will be generated daily for each working area equipped with flow and pressure metering. Through this, the localisation of inefficiencies will be possible, which addresses the literature gaps identified. Each job description will be assigned a water-consuming tool. When the employee tracking system detects a specific job description in a working area, the correlating water-consuming tool will receive water allocation.

The water usage will be determined based on the operational conditions, as represented by the derived formula discussed in Section 2.1.2. By doing so, the baselines will be dynamically adapted based on the operational conditions and employee movements, ensuring the accurate development of zero-waste baselines that can be used for accountability purposes. The accuracy of the baselines is of critical importance to ensure fairness and transparency. The process is schematically shown in Figure 3.

2.2. Step 2: Assess

Accurate dynamic zero-waste baselines for each working area are now available. The development of these baselines alone will not mitigate inefficiencies but merely serve as a tool. To address inefficiencies, responsibility allocation should be performed so that all employees are aware of their roles and responsibilities. Given the innovative nature of this approach, it is advisable to implement it in a phased manner to ensure the successful execution of the change management process.

Performance evaluations can be conducted by comparing the developed dynamic zero-waste baselines to the actual water usage for each working area. The employees responsible for the areas of concern will then need to answer for the performance as measured by the comparison to the developed baselines. The working areas’ performances should be made available to all stakeholders to ensure transparency and clear expectations.

An example of performance distribution includes platforms and reports to be distributed to all stakeholders to view their performance.

2.3. Step 3: Execute

This step includes repercussions if expectations are met or not. It is important to establish a clear procedure to guide these actions. To align with the literature, recommendations and repercussions should steer clear of social, economic, and political implications, as they often lead to conflict [42]. Instead, it is proposed that hierarchical structures are introduced, where individuals or crews failing to comply are held accountable by higher authorities, as suggested by the literature [43].

Conversely, individuals or crews exceeding expectations deserve recognition, in line with the literature, which suggests the incorporation of social pressure to encourage adherence [35,54]. This involves publicly acknowledging exceptional performance by featuring their names on widely accessible platforms or presenting them with distinctive attire as a symbol of their outstanding achievements. A typical hierarchical structure is shown in Figure 4. The hierarchical structure within a mine is critical to understanding the flow of responsibilities across departments. This structure will be used to identify the key stakeholders responsible for water management. Here, a bottom-up approach can be followed to give personnel at the lower levels of the hierarchy an opportunity to address the identified issues before being publicly identified. Doing this will lead to greater engagement and ownership across all levels.

After all stakeholders are made aware of the repercussions, opportunities to discuss performance should be established before any steps are taken.

2.4. Step 4: Communicate

The final step involves the communication of performance and repercussions. Open lines of communication should be established, such as recurring meetings, personal messages, notice boards in working areas, etc., where all problems, solutions, and timelines are discussed. All stakeholders should be present to ensure transparency.

2.5. Verification

The verification of the method should determine if the proposed method successfully addresses the identified problem. Table 1 in Section 1.3 summarises the components of an accountability method; therefore, the proposed method should adhere to all the listed components to comply with the literature. Each component is listed together with the outcomes contributing to the corresponding component. The method is deemed as successful if all the components have been addressed. The logic map for this process is displayed in Figure 5.

3. Results

This section entails the implementation of the proposed method, described in Figure 1, to a case study conducted on Mine A. Mine A is a deep-level gold mine situated in South Africa. The mining is based on a scattered mining method, together with an integrated backfill support system. This operation consists of 35 active working areas, producing approximately 6000 kg of gold annually. To produce the gold, Mine A dewaters approximately 36 mL of water per day, which is cooled by large refrigeration systems for reuse underground. The mine’s pumping and refrigeration systems account for roughly 16% and 18% of the total power consumption of Mine A, respectively.

3.1. Existing Inefficiency Detection Methods

For this paper, existing inefficiency detection techniques were implemented to serve as a benchmark for comparison with the developed method. The current inefficiency detection methods that allow for the localisation of inefficiencies include drop tests and service audits.

Service water drop tests were conducted on a South African gold mine. These drop tests were conducted during times when no personnel were working underground to identify the wastages present during non-mining periods. A typical portable flowmeter used to measure the water consumption on each half-level is seen in Figure 6. All the crosscuts were closed and then opened one by one to measure the water consumption of each crosscut when no mining was taking place. Since no personnel were working during these times, any water consumption was attributed to wastage.

Table 2. Drop test results from case study mine.

Month 1			Month 2			Month 3
Level	XC	Wastage [L/s]	Level	XC	Wastage [L/s]	Level	XC	Wastage [L/s]
A	35	44	A	51	11	B	51	17
A	57	18	B	36	10	A	61	11
A	38	15	A	32	10	A	36	11
B	33	15	A	63	8	B	36	11
A	36	14	A	38	8	A	62	9

displays the drop test results for the previous three months for the five crosscuts with the most wastage. It is clear that some areas are present each month without any mitigation, as highlighted in the table. Furthermore, these values are instantaneously measured values and only obtained monthly at times when no personnel are present underground. Therefore, no accountability can be assigned for these instantaneous results and new information is only made available each month after the drop tests are conducted. These results confirm that after three months, no work has been performed to mitigate costly inefficiencies.

In addition, service water audits were conducted for a production level in Mine A. Here, numerous inefficiencies were identified and recommended to be fixed. Three months later, a follow-up audit was conducted to determine if the previously identified inefficiencies had been rectified. The results are depicted below in Figure 7. From the results, it is clear that numerous recurring inefficiencies were detected during the follow-up audit.

Furthermore, no indicator exists to determine whether the identified wastages have been mitigated other than by conducting a follow-up underground visit. It is important to note that an audit such as this one is extremely time-intensive. The service water audit of the specific level consisted of two employees having to cover a total of approximately 22.2 km within 6.5 h, with no indication of whether the identified inefficiencies had been mitigated.

3.2. Step 1: Define

The following section will focus on the “define” step of the developed method.

3.2.1. Major Water Consumers

The primary intermittent water users at Mine A have been identified. Actual water tests were conducted to establish the water consumption for all the primary intermittent water equipment at various pressures.

The data were captured using a portable ultrasonic water flow meter and portable pressure gauge. These data were used to develop scatter plots that were utilised to establish trendline equations for all water users, as displayed below in Table 3. This process ensured that the water usage for all the equipment was accurately determined using the actual conditions. The actual water pressures, measured by meters installed in each crosscut, were input into the formula to calculate the water consumption for each piece of equipment based on the observed pressure in the working area. It should be noted that results are only shown for one working area; however, the same principle was applied to the remaining areas.

In the table below, y and x are as follows:

y—Water consumption in l/s;

x—Water pressure in bar.

Table 3. Trendline equations for intermittent water users.

Equipment Name	Trendline Equation	${\mathit{R}}^2$ Value
10 mm open hose	$y=-0.0015x^2+0.0823x+0.0127$	0.9986
25 mm open hose	$y=1.7913\ln x+0.3892$	0.9992
Venturi	$y=-0.0009x^2+0.0542x+0.0345$	0.9934
Victoria rockdrill	$y=0.2482\ln x-0.0284$	0.9907
Conventional rockdrill	$y=-0.0003x^3+0.0075x^2-0.0144+0.2681$	0.9904
Nozzle	$y=-0.0014x^2+0.0707x+0.0937$	0.9981
Omni prop	$y=0.0181x+0.076$	0.9979
Blue jackpot	$y=0.0001x^2+0.006x+0.0062$	0.9821
Waterjet	No correlation function as it uses a high-pressure pump to ensure designed water flow—3.42 L/s.

Table 3. Trendline equations for intermittent water users.

Equipment Name	Trendline Equation	${\mathit{R}}^2$ Value
10 mm open hose	$y=-0.0015x^2+0.0823x+0.0127$	0.9986
25 mm open hose	$y=1.7913\ln x+0.3892$	0.9992
Venturi	$y=-0.0009x^2+0.0542x+0.0345$	0.9934
Victoria rockdrill	$y=0.2482\ln x-0.0284$	0.9907
Conventional rockdrill	$y=-0.0003x^3+0.0075x^2-0.0144+0.2681$	0.9904
Nozzle	$y=-0.0014x^2+0.0707x+0.0937$	0.9981
Omni prop	$y=0.0181x+0.076$	0.9979
Blue jackpot	$y=0.0001x^2+0.006x+0.0062$	0.9821
Waterjet	No correlation function as it uses a high-pressure pump to ensure designed water flow—3.42 L/s.

All the equations in Table 3 have $R^2$ values close to one (as discussed in Section 2.1.2), indicating that they accurately represent the water consumption of all the water-consuming equipment and will be used accordingly. This addresses the gap identified in the existing research.

Figure 8 displays the pictures taken by one of the authors of the water-consuming equipment at the case study mine.

3.2.2. Employee Tracking Incorporation

For this study, an MPL system was used. This system is widely implemented in South African gold mines. The MPL data are stored in the mine’s internal database. It is important to note that the data stored in the database include the tracking information for all employees for all areas in the mine. However, for the purpose of this study, the job descriptions of water-consuming personnel were filtered within the specific areas in question. For this application, Python version 3.12 was used to filter the data; however, any platform can be used to sort the data into the desired format.

The first step was to determine the job descriptions that typically consume water. These are as follows:

• Night shift cleaner;
• Rock drill operator;
• Waterjet operator;
• General miner.

The developed Python script accessed the mine’s MPL database, whereafter it automatically sorted the data based on the employee activity for the given job descriptions. The specific working area was inserted into the script along with the desired date. The filtered data were stored and used to develop the zero-waste baselines. The filtered data for working area 1 are shown below in Figure 9. Here, the selected job descriptions’ activity can be seen together with the total employee movement in working area 1. By using the data shown below, the time-of-use patterns of the handheld water-consuming equipment could be determined more accurately.

3.2.3. Dynamic Zero-Waste Baseline Development

As mentioned, each job description is allocated a specific piece of water-consuming equipment. The job description and corresponding equipment are shown below in

Table 4. Equipment allocation for specific job description.

Job Description	Task	Water-Consuming Equipment
Night-shift cleaner	Cleans the face after blasting.	Rubber hoses and bazookas
Rock drill operator	Drills holes in rockface for explosions.	Victoria and conventional rock drill
Waterjet operator	Cleans the face after blasting.	Waterjet
General miner	Assists with the supporting of rock formations.	Venturi, nozzle, omni prop, and blue jackpot

.

Together with this, the amount of equipment in the working areas is obtained. Each user entering the area is allocated their specific equipment, but the data provided below are used as a secondary form of assurance for verification. This means that if the MPL data pick up that, for instance, five waterjet users are in the active area, water will not be assigned to all five users, as there are only three waterjets installed in the specific working area. Doing this helps to ensure that the developed baseline is accurate.

Table 5. Amount of equipment in working area 1.

Equipment Name	Quantity
Waterjets	3
Victoria rock drills	13
10 mm rubber hose	8
25 mm rubber hose	2
Bazookas	0
Venturi	1

displays the amount of equipment in working area 1.

Incorporating all the above-mentioned factors, the dynamic zero-waste baselines can be developed. Figure 10 shows the zero-waste baseline for working area 1.

Figure 10 displays the water wastage throughout the day, even when no water-consuming personnel were present in the area. To understand why wastage was present, an underground visit to working area 1 was conducted. Figure 11 depicts the inefficiencies observed during the underground visit. These inefficiencies are typically present throughout the day. The proposed accountability method addresses the issues seen in Figure 11 to ensure that the identified water inefficiencies are not only identified and quantified but also mitigated.

3.3. Step 2: Assess

Figure 12 displays a screenshot of the user interface for the performance platform developed to monitor and track system performance. The platform was developed for the purpose of this study; however, any platform could be used that visually represents the desired data. The platform serves as a tool for personnel to view their performance at any given time.

The platform allows the user to select the desired working area. The users have access to information, such as the actual water pressure, flow, and temperature. Together with this, the zero-waste baseline for that specific area is available. Once they have accessed the parameters, the users can select certain dates which display the data captured for that period.

The key stakeholders can stay informed through an automated daily report, specifically developed for the novel accountability approach. This report includes the same information as that seen on the platform. Here, the stakeholders can view their performance, which will be discussed during scheduled meetings. The report includes problematic areas and action lists, which will be discussed during daily meetings.

3.4. Step 3: Execute

As mentioned, personnel will need to answer to supervisors when expectations are not met. Hierarchical structures have been used to steer clear of financial incentives, which may lead to labour disputes [42]. The following hierarchical structure was put in place, as seen in Figure 13. Here, the engineering and mine managers discuss the results with the mine captains. The mine captains are in discussion with their shift bosses, who in turn appoint certain crew members to mitigate the identified inefficiencies. The engineering and mine managers give feedback to the general manager. For this paper, the focus was placed on the accountability up to the mine captains, as indicated by the red dotted lines in Figure 13.

Following discussions with the mining personnel, it was decided to implement a 3-strike point system. When an inefficiency is identified, it is included in the action list. Each time that the same inefficiency is discussed without any progress or feedback, the mine captain receives a strike. When three strikes have been received, the mine captain is called in by the general manager.

Conversely, the top-performing teams’ names are displayed on the notice board, as seen in Figure 14. Through this, a culture is developed where employees strive to earn public recognition for their good work. Furthermore, all the stakeholders work towards the same goal, which is to ensure the life-of-mine and efficient mining practices.

3.5. Step 4: Communication

Existing daily meetings are used to discuss tasks. A slot is allocated during these meetings to discuss the previous day’s water performance. The meetings are held between all mine captains and the relevant managers. Here, the mine captains provide feedback with regards to their performance on the previous day. Once a month, feedback sessions with the general manager are conducted to bring them up to speed and obtain their input if needed. This aligns with the research, which stresses that this mechanism is most effective when employees must face managers in person to account for their performance.

3.6. Method Integration and Results

By following all the above-mentioned steps of the developed accountability method, inefficiencies are identified, mitigated, and continuously monitored to ensure sustainability. The results of implementing the method are shown below. The data shown in the figures below compare the actual water consumption data received from the case study mine to the dynamic zero-waste baselines developed in this study for working area 1.

• High service water consumption is identified by using the developed dynamic zero-waste baselines, as depicted in Figure 15. Here, it is clear that service water is wasted throughout the day, even when no personnel are present in the area.
• To ensure the smooth integration of the accountability method, it was suggested that a phased approach be followed. Phase one consisted of reducing water wastages when no personnel were present underground. This encouraged employees to gradually become more conscious of reducing water wastage.
• Mine captain A was made aware of the water wastage in his area by using the developed reporting and communication structures. He was able to access his area’s performance on the dashboard.
• During the allocated slot in a daily meeting, mine captain A was made aware of the phased approach that was going to be implemented. This entailed closing the water isolation valve at the entrance to the working area when the last employee leaves the area. This cuts off the water supply to this working area, resulting in reduced water wastage during times when no personnel are present in the area. Hierarchical structures were used, and mine captain A was made aware that he would be held responsible for ensuring that the valve was closed. He, in turn, allocated a responsible person to mitigate the identified issue.
• Figure 16 displays the results from the first phase, after a week of implementation.
  Figure 16 confirms that the water valve was indeed closed during times when no personnel were observed underground. Mine captain A successfully guided his team to ensure that wastages were reduced during times when no personnel were observed. It is important to note that in Figure 16, the zero-waste baseline exceeds the actual water consumption profile from approximately 22:00. This happens because the system detects a potential water user and attributes water usage to them. However, in some rare cases, these individuals are present in the area without consuming any water. The main takeaway is that the actual consumption remains below that of the zero-waste baseline. The zero-waste baseline represents the total allowable water consumption during the day based on employee movement.
• Next, the second phase was discussed during a daily meeting. This phase included reducing water wastage throughout the day, including active working hours.
• The developed structures were used to notify mine captain A of issues identified in his area by comparing the actual water consumption to that of the developed zero-waste baseline.
• Mine captain A, after discussions with his team, mentioned that workers opened rubber hoses to release the water pressure in the main pipeline as it posed pressure problems. These rubber hoses were left open, which meant that water was wasted throughout the day. This was due to a faulty pressure relief valve (PRV) that did not reduce the water pressure at the inlet of the area as designed. Furthermore, it was mentioned that a burst manifold was observed, which should be rectified.
• In a daily meeting, it was discussed and concluded that the engineering team was responsible for fixing the faulty PRV. A fitter for the specific section was tasked to replace the broken manifold.
• Figure 17 displays the results of the second phase. Here, it can be seen that the actual water consumption profile follows that of the developed zero-waste baseline. It should be noted that wastages are still observed for certain periods of the day, which also need to be attended to. However, the results presented here confirm that by implementing accountability, water wastage can be reduced throughout the day. Following the developed method, water wastage was reduced from approximately 0.34 mL per day to 0.03 mL per day for working area 1. This equates to a total annual energy cost benefit of approximately USD 1.6 million (ZAR 28.7 million) for Mine A.

Figure 15. Water wastage identified before method implementation.

Figure 15. Water wastage identified before method implementation.

Figure 16. Results after implementing phase 1.

Figure 16. Results after implementing phase 1.

Figure 17. Results after implementing phase 2.

Figure 17. Results after implementing phase 2.

Figure 18 compares the actual water flow profiles before implementation with the results achieved after implementing both phase 1 and phase 2. Phase 1 focused on reducing water wastage during periods when no personnel were underground (between 15:00 and 21:00). After this goal was achieved, the focus was shifted to reducing water wastage throughout the day. The significant achievements are displayed below in Figure 18.

Figure 19 displays the total daily water usage over a period of 29 days (excluding weekends). The issues, as mentioned, were discussed on day 1. The team fixed the identified issues by day 7, whereafter a significant water reduction was observed. From there, the daily usage was monitored to ensure compliance. On day 20, high water usage was observed, addressed, and mitigated within one day.

Compared to the existing methods, as discussed in Section 3.1, the developed method enabled inefficiency mitigation within one week of identification. Additionally, the mitigated inefficiencies can now be monitored daily, ensuring compliance and sustainability. If inefficiencies resurface, they will be addressed as demonstrated in this paper. While some inefficiencies may be more complex and require additional time to mitigate, the developed approach ensures they will be rectified in a timely manner.

3.7. Verification

The research has identified several components that should be included in a successful accountability system. These components are listed in Table 1. The developed method was verified by comparing the identified components to the outcomes of this study to ensure that each was addressed. Each component is listed together with the corresponding step of the proposed method where the component is addressed, as seen in

Table 6. Accountability components together with the step where the component is addressed.

Component	Addressed in Step Number
Clarified expectations	1 and 3 (define and execute)
Improved performance	2 and 4 (assess and communicate)
Transparency and fairness	2, 3, and 4 (assess, execute, and communicate)
Risk management	3 and 4 (execute and communicate)
Goal alignment	1, 3, and 4 (define, execute, and communicate)
Employee engagement and motivation	3 and 4 (execute and communicate)

.

The proposed method is deemed as successful, as all the components as set out are addressed by the developed method.

4. Conclusions

Chilled-water wastage is a problem in deep-level gold mining operations. This needs to be addressed to decrease the operational costs of mining operations. Before inefficiencies can be addressed, they need to be identified. However, the existing work conducted on inefficiency detection relies on broad, infrequent, and labour-intensive work. Furthermore, no comprehensive method could be found to mitigate the identified issues and monitor the mitigated issues to ensure sustainability.

This study addresses water wastage in South African deep-level gold mines by implementing an accountability method that integrates dynamic zero-waste baselines. These baselines are constructed for each working area of a mine and represent an optimal water usage profile based on the working conditions, employee movements, and equipment utilised. The actual water consumption profile is compared to the baseline to successfully identify water wastage in closer proximity to the working areas, where most inefficiencies occur. This approach ensures real-time inefficiency detection and localised monitoring using flow and pressure meters installed in each working area.

The accountability framework developed in this study used four key components: define, assess, execute, and communicate. The expectations and goals were first outlined (define) to ensure all personnel were aware of the water usage standards. Next, daily water usage data were compared to the zero-waste baseline (assess) to identify inefficiencies. Immediate actions were taken to address the identified issues, utilising a hierarchy of accountability to assign responsibilities (execute). Daily and monthly meetings (communicate) were implemented to discuss water usage performance, reinforcing responsibility among the mine captains and their teams.

This system was implemented in phases, firstly focusing on reducing water wastage during times when no personnel were underground. Next, the focus was placed on water wastage throughout the day. By following a phased approach, this ensured that the personnel were gradually encouraged to adopt water-saving behaviours.

The proposed method reduced the daily water wastage in working area 1 from 0.34 mL to 0.03 mL, which equates to a total annual energy cost benefit of approximately USD 1.6 million (ZAR 28.7 million) for Mine A.

Furthermore, inefficiencies were not only identified faster compared to previous methods, but they were also mitigated and monitored to ensure sustainability.

The results obtained show that by implementing accountability, water wastage can be reduced throughout the day. The method is deemed successful, seeing that all the objectives set out in this paper have been met.

5. Recommendations for Future Research

This study established a single baseline for comparison against the actual water consumption of a mining operation. It is recommended that maximum and minimum ranges be defined as error intervals to create an operational water consumption range for each area, thereby compensating for numerical fluctuations.